KR20220122429A - Method, server and computer program for providing legal documents clustering search service using artificial intelligence - Google Patents

Abstract

Provided are a method, a server, and a computer program for providing a legal document clustering search service using artificial intelligence (AI). According to various embodiments of the present invention, the method for providing the legal document clustering search service using AI performed by a computing device comprises: a step of receiving at least one search word; a step of generating at least one dynamic cluster by using the at least one search word and performing dynamic clustering for a plurality of pre-stored legal documents; a step of selecting at least one static cluster related to the at least one input search word among a plurality of static clusters generated by statically clustering the plurality of pre-stored legal documents; and a step of providing a search result including the generated at least one dynamic cluster and the selected at least one static cluster. The present invention has an advantage of providing more accurate search results.

Description

Method of providing legal document clustering search service using artificial intelligence, server and computer program

Various embodiments of the present invention relate to a method of providing a legal document clustering search service using artificial intelligence, a server, and a computer program.

In general, since legal documents have the characteristics of being interpreted legally by applying them to issues of law, these legal documents help people to establish desirable legal standards for each issue they face.

Here, the legal document means a contract or judgment written in legal language. More than 1.5 million legal documents are created every year, and there are many legal documents for each issue.

Conventionally, a legal document search system has been provided so that the general public can conveniently view and access these legal documents. have. In particular, among legal documents, court precedents include issues and laws, but there is no keyword specifically specifying the issues, so it is difficult to collect legal documents on specific issues.

In addition, less than 0.3% of the legal documents currently publicly available are not digitized and most of them are not digitized and exist in the form of paper or images, so there is a problem that it is not easy for users to search.

In addition, in the case of the conventional legal document search system, since it is based on a keyword search within a document, there is a disadvantage that even unrelated legal documents are searched for a simple keyword search. In order to do this, there is a problem that all documents in the search results must be read one by one.

The problem to be solved by the present invention is to overcome the disadvantages and problems of the conventional legal document search system as described above, in response to receiving a search word from a user, a dynamic cluster generated by dynamically clustering legal documents according to the search word. Artificial intelligence that can provide more accurate search results by providing search results including static clusters related to the relevant search term among a plurality of static clusters generated by performing static clustering of a plurality of legal documents based on legal document information It is to provide a method of providing a search service for clustering legal documents using intelligence, a server and a computer program.

The problems to be solved by the present invention are not limited to the problems mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the following description.

A method of providing a legal document clustering search service using artificial intelligence according to an embodiment of the present invention for solving the above-described problems, in a method performed by a computing device, includes the steps of: receiving one or more search terms; generating one or more dynamic clusters by performing dynamic clustering on a plurality of pre-stored legal documents using the above search words; The method may include selecting one or more static clusters related to one or more search terms and providing a search result including the one or more generated dynamic clusters and the selected one or more static clusters.

In various embodiments, the generating of the one or more dynamic clusters may include selecting a plurality of legal documents including the input one or more search terms from among the plurality of pre-stored legal documents, and vectorizing the selected plurality of legal documents. and calculating a similarity between the plurality of vectorized legal documents, and performing dynamic clustering on the selected plurality of legal documents based on the calculated similarity.

In various embodiments, the selecting of a plurality of legal documents including the inputted one or more search words includes selecting a legal document including the inputted one or more search terms from among the plurality of pre-stored legal documents at least once. , when the number of the selected legal documents exceeds a preset maximum value, the input of one or more search terms for the plurality of pre-stored legal documents until the number of the selected legal documents is less than or equal to the preset maximum value It may include the step of sequentially increasing the minimum number of inclusions.

In various embodiments, the generating of the one or more dynamic clusters includes selecting a legal document including all of the plurality of search terms from among the plurality of pre-stored legal documents when a plurality of search words are input, and the selected legal document Dynamic clustering is performed for , and when the selected legal document is less than a preset minimum value, priority among the plurality of search terms until the number of the selected legal documents is equal to or greater than the preset minimum value - The priority is the plurality of The method may include sequentially excluding search words with a lower value - which is set based on at least one of an input order of search words and an operator input together with the plurality of search words.

In various embodiments, the method further comprises generating a plurality of static clusters by performing static clustering on the plurality of pre-stored legal documents, wherein the generating of the plurality of static clusters includes: the plurality of pre-stored laws A plurality of first static clusters are formed by performing static clustering according to each document as a reference document - the reference document includes at least one of precedents, statutes, documents, and comments referenced by each of the plurality of pre-stored legal documents generating, by performing static clustering of each of the plurality of pre-stored legal documents according to a related law, wherein the related law includes at least one of laws and articles cited by each of the plurality of pre-stored legal documents. generating a second static cluster of , and generating a plurality of third static clusters by performing static clustering of each of the plurality of pre-stored legal documents according to an issue dealt with by each of the plurality of pre-stored legal documents. may include

In various embodiments, the selecting of the one or more static clusters may include calculating a relevance between the plurality of static clusters and the input one or more search terms, and selecting one or more static clusters in which the calculated relevance is equal to or greater than a preset value. may include steps.

In various embodiments, the selecting of the one or more static clusters having the calculated similarity equal to or greater than a preset value may include, when a plurality of search words are input, one or more of the plurality of static clusters having relevance to all search terms greater than or equal to the preset value. The method may include selecting one or more static clusters or selecting one or more static clusters in which an average value of relevance for each of the plurality of search terms is equal to or greater than the preset value.

In various embodiments, the providing of the search result may include, when the number of dynamic clusters generated using a first search term and static clusters selected using the first search term is less than a preset number, synonyms of the first search term and providing a dynamic cluster generated using a second search word or a third search word that is a similar word to the first search word and a static cluster selected using the second search word or the third search word.

In various embodiments, the receiving of the one or more search terms includes providing a plurality of query contents to the user in response to receiving a request for provision of a legal document search service from the user, the plurality of queries provided by the user It may include the steps of receiving a plurality of answers to the content, extracting one or more keywords by analyzing the plurality of input answers, and selecting the extracted one or more keywords as a search word.

In various embodiments, the method further comprises collecting a plurality of legal documents and processing and storing the plurality of collected legal documents, wherein the processing and storing of the plurality of collected legal documents includes: When the legal document of When the language of the plurality of legal documents is a foreign language, the method may include automatically translating the foreign language into a preset language, and processing and storing the automatically translated legal document on the preset template.

A legal document clustering search service providing server using artificial intelligence according to another embodiment of the present invention for solving the above problems is loaded into a processor, a network interface, a memory and the memory, and is executed by the processor A computer program comprising: an instruction for receiving one or more search words an instruction for generating, an instruction for selecting one or more static clusters related to the one or more inputted search words from among a plurality of static clusters generated by static clustering the plurality of pre-stored legal documents, and the generated one or more dynamic clusters and the selected one An instruction for providing a search result including the above static clusters may be included.

A computer program recorded on a computer-readable recording medium according to another embodiment of the present invention for solving the above-described problems is coupled to a computing device, the step of receiving one or more search words, the input one or more search words generating one or more dynamic clusters by performing dynamic clustering on a plurality of pre-stored legal documents using It may be stored in a computer-readable recording medium to execute the steps of selecting one or more static clusters related to the one or more dynamic clusters and providing a search result including the one or more generated dynamic clusters and the selected one or more static clusters.

Other specific details of the invention are included in the detailed description and drawings.

According to various embodiments of the present invention, in response to receiving a search word from a user, static clustering of a plurality of legal documents is performed based on the dynamic cluster generated by dynamically clustering legal documents according to the search word and legal document information. By providing a search result including a static cluster related to a corresponding search term among a plurality of generated static clusters, there is an advantage in that it is possible to overcome the disadvantages and problems of the conventional legal document search system as described above and provide more accurate search results. .

Effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the following description.

1 is a diagram illustrating a legal document clustering search service providing system using artificial intelligence according to an embodiment of the present invention.
2 is a hardware configuration diagram of a legal document clustering search service providing server using artificial intelligence according to another embodiment of the present invention.
3 is a flowchart of a method for providing a legal document clustering search service using artificial intelligence according to another embodiment of the present invention.
4 is a flowchart illustrating a method of dynamically clustering legal documents using a search word input by a user, according to various embodiments.
5 is a flowchart illustrating a method of selecting a static cluster corresponding to a search term from among a plurality of static clusters generated by performing static clustering on a plurality of legal documents, according to various embodiments.
6 is a view exemplarily illustrating a process of collecting and storing legal documents applicable to various embodiments.
7 is a diagram exemplarily illustrating a legal document clustering search service providing user interface (UI) applicable to various embodiments.

Advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only these embodiments allow the disclosure of the present invention to be complete, and those of ordinary skill in the art to which the present invention pertains. It is provided to fully inform those skilled in the art of the scope of the present invention, and the present invention is only defined by the scope of the claims.

The terminology used herein is for the purpose of describing the embodiments and is not intended to limit the present invention. In this specification, the singular also includes the plural, unless specifically stated otherwise in the phrase. As used herein, “comprises” and/or “comprising” does not exclude the presence or addition of one or more other components in addition to the stated components. Like reference numerals refer to like elements throughout, and "and/or" includes each and every combination of one or more of the recited elements. Although "first", "second", etc. are used to describe various elements, these elements are not limited by these terms, of course. These terms are only used to distinguish one component from another. Therefore, it goes without saying that the first component mentioned below may be the second component within the spirit of the present invention.

Unless otherwise defined, all terms (including technical and scientific terms) used herein will have the meaning commonly understood by those of ordinary skill in the art to which this invention belongs. In addition, terms defined in a commonly used dictionary are not to be interpreted ideally or excessively unless clearly specifically defined.

As used herein, the term “unit” or “module” refers to a hardware component such as software, FPGA, or ASIC, and “unit” or “module” performs certain roles. However, “part” or “module” is not meant to be limited to software or hardware. A “part” or “module” may be configured to reside on an addressable storage medium or may be configured to reproduce one or more processors. Thus, by way of example, “part” or “module” refers to components such as software components, object-oriented software components, class components and task components, processes, functions, properties, Includes procedures, subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays, and variables. Components and functionality provided within “parts” or “modules” may be combined into a smaller number of components and “parts” or “modules” or additional components and “parts” or “modules”. can be further separated.

Spatially relative terms "below", "beneath", "lower", "above", "upper", etc. It can be used to easily describe the correlation between a component and other components. A spatially relative term should be understood as a term that includes different directions of components during use or operation in addition to the directions shown in the drawings. For example, when a component shown in the drawing is turned over, a component described as “beneath” or “beneath” of another component may be placed “above” of the other component. can Accordingly, the exemplary term “below” may include both directions below and above. Components may also be oriented in other orientations, and thus spatially relative terms may be interpreted according to orientation.

In this specification, a computer means all types of hardware devices including at least one processor, and may be understood as encompassing software configurations operating in the corresponding hardware device according to embodiments. For example, a computer may be understood to include, but is not limited to, smart phones, tablet PCs, desktops, notebooks, and user clients and applications running on each device.

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

Each step described in this specification is described as being performed by a computer, but the subject of each step is not limited thereto, and at least a portion of each step may be performed in different devices according to embodiments.

In this specification, the method for providing a legal document clustering search service using artificial intelligence described in this specification is described as a method of classifying legal documents and providing a legal document corresponding to a search word input by a user, but this is an example However, the present invention is not limited thereto, and it can be applied to documents other than legal documents.

1 is a diagram illustrating a legal document clustering search service providing system using artificial intelligence according to an embodiment of the present invention.

Referring to FIG. 1 , a legal document clustering search service providing system using artificial intelligence according to an embodiment of the present invention includes a legal document clustering search service providing server 100 , a user terminal 200 and an external server 300 . can do.

Here, the legal document clustering search service providing system using artificial intelligence shown in FIG. 1 is according to an embodiment, and its components are not limited to the embodiment shown in FIG. 1, and may be added, changed or may be deleted.

In one embodiment, the legal document clustering search service providing server 100 (hereinafter, “server 100”) receives a search term from a user and provides a legal document clustering search service that provides search results related to the inputted search term. can do.

For example, the server 100 may be connected to the user terminal 200 through the network 400 , and a UI (eg, 10 in FIG. 7 ) for using the legal document clustering search service to the user terminal 200 . may be provided, and a search word may be input through the UI 10 and a search result may be output.

In various embodiments, the server 100 may provide a web or application-based legal document clustering search service. However, the present invention is not limited thereto.

Here, the user terminal 200 may include a display on at least a portion of the user terminal 200 , and may include an operating system for driving an application-based legal document clustering search service provided from the server 100 . For example, the user terminal 200 may be a smartphone (Smart-phone), but is not limited thereto, and the user terminal 200 is a wireless communication device that guarantees portability and mobility, and includes navigation and personal communication (PCS). System), GSM (Global System for Mobile communications), PDC (Personal Digital Cellular), PHS (Personal Handyphone System), PDA (Personal Digital Assistant), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)- 2000, W-CDMA (W-Code Division Multiple Access), Wibro (Wireless Broadband Internet) terminal, smartpad, tablet PC, etc. all kinds of handheld-based wireless communication devices. may include

Also, here, the network 400 may refer to a connection structure capable of exchanging information between respective nodes such as a plurality of terminals and servers. For example, the network 400 includes a local area network (LAN), a wide area network (WAN), the Internet (WWW), a wired/wireless data communication network, a telephone network, a wired/wireless television communication network, and the like. do.

In addition, here, the wireless data communication network is 3G, 4G, 5G, 3rd Generation Partnership Project (3GPP), 5th Generation Partnership Project (5GPP), Long Term Evolution (LTE), World Interoperability for Microwave Access (WIMAX), Wi-Fi (Wi- Fi), Internet, LAN (Local Area Network), Wireless LAN (Wireless Local Area Network), WAN (Wide Area Network), PAN (Personal Area Network), RF (Radio Frequency), Bluetooth network, Near-Field Communication (NFC) networks, satellite broadcast networks, analog broadcast networks, Digital Multimedia Broadcasting (DMB) networks, and the like are included, but are not limited thereto.

In various embodiments, the server 100 may provide an artificial intelligence-based legal document clustering search service. For example, the server 100 may cluster a plurality of pre-stored legal documents by using a pre-learned artificial intelligence model, and may provide a result derived therefrom to the user as a search result.

Here, the artificial intelligence model is a model learned by machine learning technology such as an unsupervised learning method, and may be a model that performs clustering on a plurality of legal documents. For example, AI models include K-Means model (or K-Means++ model), Mean Shift model, Gaussian Mixture Model (GMM), and Density Based clustering. Spatial Clustering of Applications with Noise, DBSCAN) model, Self-Organizing Map (SOM) model, and any one of ISODATA model may be used, but is not limited thereto, and two or more of the above models are used (eg: ensemble) to cluster legal documents.

In one embodiment, the external server 300 may be connected to the server 100 through the network 400, and various information/data ( Example: A plurality of legal documents, a plurality of static clusters generated by static clustering of a plurality of legal documents, search history for a plurality of users, etc.) can be stored and managed, and the server 100 can store and manage legal documents using artificial intelligence. It is possible to store and manage various types of information/data (eg, results of performing dynamic clustering according to a search word input by a user, etc.) generated by providing the clustering search service.

In various embodiments, the external server 300 may be a storage server (eg, a database server) separately provided outside the server 100 , but is not limited thereto, and the external server 300 is used to store various types of information and data. ) can be stored and managed in a separate space in the server 100 without having to separately provide it outside the server 100 , and store and manage the database in the corresponding space. Hereinafter, with reference to FIG. 2 , a hardware configuration of the server 100 providing a legal document clustering search service using artificial intelligence will be described.

2 is a hardware configuration diagram of a legal document clustering search service providing server using artificial intelligence according to another embodiment of the present invention.

2, the server 100 according to another embodiment of the present invention is one or more processors 110, a memory 120 for loading a computer program 151 executed by the processor 110, It may include a bus 130 , a communication interface 140 , and a storage 150 for storing a computer program 151 . Here, only the components related to the embodiment of the present invention are shown in FIG. 2 . Accordingly, those skilled in the art to which the present invention pertains can see that other general-purpose components other than the components shown in FIG. 2 may be further included.

The processor 110 controls the overall operation of each component of the server 100 . The processor 110 includes a central processing unit (CPU), a micro processor unit (MPU), a micro controller unit (MCU), a graphic processing unit (GPU), or any type of processor well known in the art. can be

In addition, the processor 110 may perform an operation for at least one application or program for executing the method according to the embodiments of the present invention, and the server 100 may include one or more processors.

In various embodiments, the processor 110 temporarily and/or permanently stores a signal (or data) processed inside the processor 110 . Random Access Memory (RAM) and ROM (Read -Only Memory, not shown) may be further included. In addition, the processor 110 may be implemented in the form of a system on chip (SoC) including at least one of a graphic processing unit, a RAM, and a ROM.

The memory 120 stores various data, commands and/or information. The memory 120 may load the computer program 151 from the storage 150 to execute methods/operations according to various embodiments of the present disclosure. When the computer program 151 is loaded into the memory 120 , the processor 110 may perform the method/operation by executing one or more instructions constituting the computer program 151 . The memory 120 may be implemented as a volatile memory such as RAM, but the technical scope of the present disclosure is not limited thereto.

The bus 130 provides a communication function between the components of the server 100 . The bus 130 may be implemented as various types of buses, such as an address bus, a data bus, and a control bus.

The communication interface 140 supports wired/wireless Internet communication of the server 100 . Also, the communication interface 140 may support various communication methods other than Internet communication. To this end, the communication interface 140 may be configured to include a communication module well known in the technical field of the present invention. In some embodiments, the communication interface 140 may be omitted.

The storage 150 may non-temporarily store the computer program 151 . When performing the precision diagnosis module suggestion process through the server 100 , the storage 150 may store various types of information necessary to provide .

The storage 150 is a non-volatile memory such as a read only memory (ROM), an erasable programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM), a flash memory, a hard disk, a removable disk, or well in the art to which the present invention pertains. It may be configured to include any known computer-readable recording medium.

The computer program 151 may include one or more instructions that, when loaded into the memory 120 , cause the processor 110 to perform methods/operations according to various embodiments of the present invention. That is, the processor 110 may perform the method/operation according to various embodiments of the present invention by executing the one or more instructions.

In one embodiment, the computer program 151 includes the steps of receiving one or more search terms, performing dynamic clustering on a plurality of pre-stored legal documents using the input one or more search words, thereby generating one or more dynamic clusters; Selecting one or more static clusters related to one or more inputted search terms from among a plurality of static clusters generated by static clustering a plurality of stored legal documents, and a search result including the generated one or more dynamic clusters and the selected one or more static clusters It may include one or more instructions for performing a method of providing a legal document clustering search service using artificial intelligence including the step of providing.

The steps of the method or algorithm described in relation to the embodiment of the present invention may be implemented directly in hardware, implemented as a software module executed by hardware, or implemented by a combination thereof. A software module may include random access memory (RAM), read only memory (ROM), erasable programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, hard disk, removable disk, CD-ROM, or It may reside in any type of computer-readable recording medium well known in the art to which the present invention pertains.

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 components, and similarly, embodiments may include various algorithms implemented as data structures, processes, routines, or combinations of other programming constructs, including C, C++ , Java, assembler, etc. may be implemented in a programming or scripting language. Functional aspects may be implemented in an algorithm running on one or more processors.

3 is a flowchart of a method for providing a legal document clustering search service using artificial intelligence according to another embodiment of the present invention.

First, the server 100 may collect and store a plurality of legal documents in order to provide a legal document clustering search service using artificial intelligence.

In various embodiments, the server 100 stores and manages a plurality of legal documents from an external server 300 (eg, the Fair Trade Commission server, the Financial Services Commission server, the Tax Tribunal server, etc.) stored in the external server 300 . of legal documents can be provided, and a plurality of provided legal documents can be processed and stored according to a preset template.

Here, most of the legal documents are not digitized, but exist in the form of paper or images. As shown in FIG. 6 , the server 100 is a legal document in the form of a plurality of images (legal documents or paper pre-stored in the form of images) legal document images, etc.) generated by shooting legal documents in the form of legal documents) can be collected, and one or more strings can be extracted by analyzing the legal documents in the form of collected images (eg OCR analysis), and the extracted one or more strings can be processed (eg, corrected for spelling and newlines, etc.) and stored on a preset template.

In other words, it is possible to digitize and store legal documents faster (within 1 minute) and more conveniently just by uploading images of legal documents without the need for a person to directly input the contents of the legal documents. Documents can be quickly digitized and secured.

In various embodiments, the server 100 collects legal documents written in a foreign language (eg, English, Japanese, Chinese, etc.) other than a preset language (eg, Korean) from the outside, the foreign language included in the collected legal documents can be automatically translated (machine translation) into a preset language, and the automatically translated legal document can be processed and stored according to a preset template.

Here, a method of translating a legal document written in a foreign language into another language is known as a method using a machine learning-based artificial intelligence model, and various techniques are known, so a specific methodology for automatic translation is not described.

In this case, the server 100 matches and stores the automatically translated legal document with the corresponding legal document, so that when the user later wants to read the legal document, the translated content and the original text can be selectively viewed. Hereinafter, a method of providing a legal document clustering search service using artificial intelligence will be described.

Referring to FIG. 3 , in step S110 , the server 100 may receive a search word from the user. For example, the server 100 may be connected to the user terminal 200 through the network 400 to provide a legal document clustering search service providing UI (10 in FIG. 7 ), and search terms through the UI 10 can be input.

Here, the search word input by the user may be a single word, but is not limited thereto, and may be in the form of a plurality of words, a phrase, a clause, or a sentence including two or more words.

In various embodiments, the server 100 may provide a plurality of query contents through the UI 10 , may receive a plurality of answers to the plurality of query contents from the user, and may receive a plurality of inputted answer contents. It is possible to provide a Q&A type search word input service in which one or more keywords are extracted by analyzing , and the extracted one or more keywords are selected as a search word.

If the user using the legal document clustering search service provided by the server 100 is a legal person, he or she can easily enter a search term to search for the legal document he or she wants. In order to do this, you may not be able to perform a proper search because you do not know what type of search term you need to enter, or you may have to go through a process such as repeating the search several times to find the legal document you want.

In order to improve this point, when the user is a non-legal, the server 100 provides a legal document that the user wants to search, that is, a query to understand the user's needs, and analyzes the answer input in response. Thus, there is an advantage that non-legal practitioners can more easily and accurately receive search results by automatically selecting a search term for searching for a legal document desired by the user.

In various embodiments, the server 100 uses pre-stored daily term data for each legal term to determine that the search term input by the user corresponds to a daily term (eg, a term used in daily conversations not used in legal documents). If it is determined, the search term for everyday terms input by the user may be automatically replaced with a search term for legal terms.

In various embodiments, when a plurality of search words are simultaneously input by the user, the server 100 may set a priority for each of the plurality of input search words.

For example, when a first search word, a second search word, and a third search word are sequentially input by the user, the server 100 assigns the first search word to the first priority, the second search word to the second priority, and The third search word may be set as the third priority.

In addition, the server 100 may receive an operator (eg, #N (here, N is a natural number indicating a priority)) for setting priorities together with a plurality of search words from the user, and based on the input operator, You can set the priority for each search term of

In various embodiments, when receiving a plurality of search words from a user, the server 100 may determine the relevance of each of the plurality of search words and group the plurality of search words based on the determined relevance. Thereafter, the server 100 does not perform an operation of providing a search result corresponding to a plurality of search terms (eg, steps S120 to S140 to be described later), but individually searches for each group of search terms grouped based on the relevance. You can perform an action that provides a result.

For example, when the server 100 classifies a plurality of search words into a first search word group and a second search word group based on the relevance of each of the plurality of search words, the server 100 goes through steps S120 to S140 to be described later to the first search word group. A search result for , and a search result for the second search word group may be individually derived and provided to the user.

Here, the method for determining the relevance between a plurality of search terms is based on preset synonyms, similar words, and antonyms data for each search word, and a plurality of users' search word input histories, etc., scoring (eg, giving a score according to the number of times searched together, etc.), and scoring A method of judging based on a result may be applied, but the present invention is not limited thereto, and other general-purpose techniques for judging a correlation between two objects may be applied.

In various embodiments, the server 100 may receive an operator (eg, parentheses) for grouping one or more search words together with a plurality of search words from the user, and group the plurality of search words based on the input operator. . However, the present invention is not limited thereto.

In step S120 , the server 100 may generate one or more dynamic clusters by performing dynamic clustering on a plurality of pre-stored legal documents using a search word input by the user.

Here, dynamic clustering may refer to a clustering method having an environment in which a size of a cluster or member nodes change according to a cluster configuration, as opposed to static clustering. Hereinafter, a dynamic clustering method performed by the server 100 will be described with reference to FIG. 4 .

4 is a flowchart illustrating a method of dynamically clustering legal documents using a search word input by a user, according to various embodiments.

Referring to FIG. 4 , in step S210 , the server 100 may select a plurality of legal documents including a search word input by the user from among a plurality of pre-stored legal documents. For example, the server 100 may select a legal document including a search word input by the user at least once from among a plurality of pre-stored legal documents.

At this time, when the selected legal document, that is, a legal document including a search word inputted from the user at least once exceeds a preset maximum value, the server 100 determines that the number of selected legal documents is less than or equal to a preset maximum value Up to , the minimum number of inclusions of the search word input by the user may be sequentially increased.

For example, when the legal document including the search word inputted from the user at least once exceeds the preset maximum value, the server 100 increases the minimum number of times to include the search word inputted from the user at least twice. You can select the legal documents included above. In addition, when the number of legal documents in which the search word input by the user is included at least twice also exceeds the preset maximum value, the server 100 counts the minimum number of inclusions 3 times or 4 times until it becomes less than or equal to the preset maximum value. It can be increased to 5 times and 5 times.

When the user enters a rather general search term, a fairly large amount of legal documents including such a search term can be searched, and the legal documents searched in this way contain a relatively large number of noise (legal documents far from the legal documents desired by the user). ), there is a problem in that it is difficult to provide accurate search results.

In order to improve this point, the server 100 may set the maximum value of legal documents to be searched in advance, and minimize noise by allowing legal documents to be selected below the maximum value by using the minimum number of included keywords. This can provide accurate results.

In various embodiments, when a plurality of search words are input by the user, the server 100 may select a legal document including all of the plurality of search words input by the user from among a plurality of pre-stored legal documents.

At this time, when the legal documents including all of the plurality of search terms input by the user are less than (or not in the case of) the preset minimum value, the server 100 processes the plurality of search terms until the number of selected legal documents is equal to or greater than the preset minimum value. You can exclude one by one.

For example, when the server 100 receives first to fifth search terms from the user, but the number of legal documents including all of the first to fifth search terms is less than a preset minimum value, the first to fifth search terms are excluded and the first It is possible to select a legal document including all of the to fourth search terms. In addition, when the number of legal documents including all of the first to fifth search terms is also less than the preset minimum value, the server 100 determines whether the number of selected legal documents is equal to or greater than the preset minimum value, until the number of the selected legal documents is equal to or greater than the preset minimum value. and sequentially excluding the second search word, it is possible to select a legal document.

Here, the server 100 may sequentially exclude a search word having a lower priority based on a priority set for each of the plurality of search words, but is not limited thereto.

When a user inputs a rather large number of search terms in order to accurately search for a desired legal document, there is a problem in that legal documents including the search term are not searched or only a very small amount of legal documents can be searched. In order to improve this problem, the server 100 may set a minimum value of legal documents to be searched in advance, and adjust the number of search terms for selecting legal documents so that legal documents can be selected more than the minimum value, so that the user You can provide the desired search results.

In step S220, the server 100 may vectorize the plurality of legal documents selected in step S210. Here, a method for vectorizing a plurality of legal documents includes a method using a word2vec algorithm, a method of vectorizing features (words included in a legal document) using a Bag Of Words (BOW) method, etc. As the technology can be applied, specific methodologies are not described.

In step S230, the server 100 may calculate a degree of similarity between the plurality of legal documents vectorized through step S220. Here, the method of calculating the similarity between the plurality of vectorized legal documents is by using the Euclidean distance, the Mahalanobis distance, and the Minkowski distance between the plurality of vectorized legal documents. Various techniques, such as a method of calculating a similarity or a method of calculating a cosine simliarity between a plurality of vectorized legal documents, may be applied, but the present invention is not limited thereto.

In step S240 , the server 100 may perform dynamic clustering on the plurality of vectorized legal documents based on the degree of similarity between the plurality of vectorized legal documents calculated in step S230 .

In various embodiments, the server 100 may determine the relevance of two or more search terms when two or more search terms are input from the user, and when it is determined that the two or more search terms are uncorrelated search terms, two or more search terms A dynamic clustering operation (eg, steps S210 to S240) for legal documents including each may be individually performed.

In various embodiments, when a plurality of search words are input from a user and the plurality of search words are classified into two or more search word groups according to the relevance of each of the plurality of search words, the server 100 performs a dynamic clustering operation for each of the two or more search word groups (eg, steps S220 to S240) may be performed individually.

Again, referring to FIG. 3 , in step S130 , the server 100 may select one or more static clusters related to one or more search terms input by the user from among a plurality of static clusters generated by static clustering a plurality of pre-stored legal documents. have.

Here, static clustering may mean that member nodes and cluster headers repeat header re-election and data transmission processes in a fixed cluster size. Hereinafter, a static cluster selection method performed by the server 100 will be described with reference to FIG. 5 .

5 is a flowchart illustrating a method of selecting a static cluster corresponding to a search term from among a plurality of static clusters generated by performing static clustering on a plurality of legal documents, according to various embodiments.

Referring to FIG. 5 , in step S310 , the server 100 may generate a plurality of first static clusters by performing static clustering of each of a plurality of pre-stored legal documents according to a reference document. For example, the server 100 performs static clustering on a plurality of pre-stored legal documents based on each precedent, statute, document, and commentary referenced by each of the plurality of pre-stored legal documents, thereby performing a plurality of first You can create static clusters.

In various embodiments, the server 100 vectorizes information on precedents, statutes, documents and comments that are referenced by a plurality of pre-stored legal documents, and calculates a degree of similarity between the plurality of legal documents based on the vectorized numerical value, A first static cluster may be generated between legal documents having a similarity greater than or equal to a preset value (eg, 90%). However, the present invention is not limited thereto.

In step S320 , the server 100 may generate a plurality of second static clusters by performing static clustering of each of a plurality of pre-stored legal documents according to related laws. For example, the server 100 generates a plurality of second static clusters by performing static clustering on a plurality of pre-stored legal documents based on each of the statutes and articles cited by each of the plurality of pre-stored legal documents. can do.

In various embodiments, the server 100 vectorizes information on statutes and articles cited by each of a plurality of pre-stored legal documents, calculates a degree of similarity between the plurality of legal documents based on the vectorized numerical value, and the degree of similarity is A second static cluster can be created between legal documents that are greater than or equal to a set value (eg, 90%). However, the present invention is not limited thereto.

In step S330 , the server 100 may generate a plurality of third static clusters by performing static clustering of each of the plurality of pre-stored legal documents according to an issue dealt with by each of the plurality of pre-stored legal documents.

In various embodiments, the server 100 vectorizes information on an issue dealt with by each of a plurality of pre-stored legal documents, calculates a degree of similarity between the plurality of legal documents based on the vectorized numerical value, and sets the similarity to a preset value ( Example: A third static cluster can be created among legal documents with more than 90%). However, the present invention is not limited thereto.

Here, the method of performing static clustering for a plurality of pre-stored legal documents performed by the server 100 in steps S310 to S330 is a method of performing the dynamic clustering (eg, steps S210 to S240 in FIG. 3 ). and may be implemented in the same or similar form, but is not limited thereto.

In various embodiments, a plurality of pre-stored legal documents may refer to a plurality of reference documents, cite a plurality of statutes and articles, or deal with a plurality of issues, so that one legal document may be assigned to several static clusters. As possible, the server 100 clusters each of a plurality of pre-stored legal documents according to reference documents, related laws and issues, and uses a soft clustering (eg, fuzzy clustering) method to cluster a plurality of pre-stored laws. You can perform clustering for each document.

In step S340 , the server 100 calculates the correlation between the plurality of first static clusters, the plurality of second static clusters, and the plurality of third static clusters generated through the steps S310 to S330 and the search word input by the user. can do. Here, the method of calculating the relevance between the plurality of static clusters and the search word may be implemented in the same or similar form as the method of calculating the degree of similarity between the plurality of vectorized legal documents (eg, step S230), but is not limited thereto. .

In step S350 , the server 100 calculates the relevance calculated based on the relevance between the plurality of first static clusters, the plurality of second static clusters, and the plurality of third static clusters calculated through step S340 and the search word input by the user. One or more static clusters that are greater than or equal to this preset value (eg, greater than 90%) can be selected. Here, a method of calculating the correlation between the search word and the static cluster may use a method of calculating the cosine similarity, but is not limited thereto.

Here, if only a static cluster having a very high relevance to a search term input by a user is selected among a plurality of static clusters, a result suitable for a search term input by the user (low noise) can be derived, but if the criterion is set too high, The problem is that there may or may not be very few static clusters selected depending on (eg, when the wrong search term is entered).

In consideration of this point, the server 100 may flexibly set a criterion (reference relevance value) for selecting a static cluster corresponding to a search word input by a user. For example, the server 100 may set the reference relevance value to a first value when the user is a non-legal, and may set the reference relevance value to a second value greater than the first value when the user is a legal person. However, the present invention is not limited thereto, and the server 100 may directly receive a reference relevance value from the user through the UI 10 , and adjust the reference relevance value accordingly, so that the user can directly set the search result while viewing the search result. can make it

In various embodiments, when a plurality of search terms are input by the user, the server 100 may select one or more static clusters in which relevance to all search terms is greater than or equal to a preset value among the plurality of static clusters.

In various embodiments, when a plurality of search words are input by the user, the server 100 may select one or more static clusters in which the average value of relevance for each of the plurality of search words is equal to or greater than a preset value. However, the present invention is not limited thereto.

Referring again to FIG. 3 , in step S140 , the server 100 may provide the user with a search result including the dynamic cluster generated in step S120 and the static cluster selected in step S130 to the user.

In various embodiments, the server 100 may provide an Excel-like view to the user when it is determined that the user is a legal person (eg, based on a check item checked for a legal person or a non-legal person at the time of membership registration). have.

At this time, the server 100 categorizes the generated search results by category (eg, precedents (special precedents, partial precedents), literature, decision cases (Tax Tribunal, Fair Trade Commission), and authoritative interpretation (reply to inquiries from the Tax Tribunal, Financial Services Commission) , Courts (Constitutional Court, Seoul High Court, Supreme Court, District Court, Seoul Administrative Court, Seoul Central District Court, High Court, Military Court), Type of Case (Constitutional Review Type Constitutional Complaint, Rights Relief Constitutional Complaint, Administrative, Civil, Criminal, unconstitutional law adjudication, war constitutional adjudication-constitutional review-type constitutional complaint, constitutional review-type constitutional complaint-rights relief constitutional complaint, authority dispute adjudication, application), date (last 1 year, 3 years, 5 years, set by the user) period), etc.) can be provided.

In addition, the server 100 may output a legal document included in the search result by applying a separate effect (eg, highlighting) to the content corresponding to the search word input by the user.

In various embodiments, when the number of dynamic clusters generated using the first search word and the number of static clusters selected using the first search word is less than a preset number, the server 100 is A dynamic cluster generated using a third search word that is a similar word to the first search word and a static cluster selected using the second search word or the third search word may be provided together.

In this case, in the legal document included in the search result, the server 100 may set different effects from the first search word to a second search word that is a synonym of the first search word and a third search word that is a similar word to the first search word. For example, the server 100 sets each of the first search word, the second search word, and the third search word to be highlighted, but sets the first search word input by the user to be highlighted in a first color, and the second search word and the third search word may be set to be highlighted with a second color and a third color different from the first color, respectively.

In various embodiments, the server 100 may provide the user with information on the recommended search word together with the search result. For example, in response to receiving a first search word from the user, the server 100 provides a dynamic cluster generated using the first search word and a static cluster selected using the first search word, but A second search word that is a synonym and a third search word that is a similar word to the first search word may be guided as a recommended search word.

In addition, the server 100 analyzes search history data for a plurality of users who use a legal document clustering search service using artificial intelligence, and one or more recommended search words (eg, to a plurality of users) related to the first search word input from the user. By analyzing the search history for the search term, it is possible to select another search term that is additionally searched for after the first search term), and guide the selected recommended search term.

In various embodiments, the server 100 responds to receiving a search word from a user, and provides information on legal documents included in the dynamic cluster generated using the search word and legal documents included in the static cluster selected using the search word. A search result including information may be provided, but the search result may be summarized and outputted on the screen.

In various embodiments, the server 100 responds to receiving a search word from a user, and provides information on legal documents included in the dynamic cluster generated using the search word and legal documents included in the static cluster selected using the search word. A search result including information may be provided, but only partial contents of each of a plurality of legal documents included in the search result may be extracted and displayed on the screen.

Here, the content extracted for each of the plurality of legal documents may be the content of a paragraph containing a lot of search words input by the user, or content included in the part (item) inputted in advance by the user, but is not limited thereto, and the user The legal document clustering search service provision UI is displayed in the legal document review usage history (e.g., which part of the legal document is focused on (counting the time the screen stays), which part is extracted and quoted a lot, etc.) can be determined based on

That is, when a large amount of legal documents are provided to the user, it is difficult to see the relevant legal documents at a glance, so only the core contents of the legal documents can be summarized and provided, or only some contents frequently checked by the user can be extracted and printed. .

The method of providing a legal document clustering search service using the aforementioned artificial intelligence has been described with reference to the flowchart shown in the drawings. For a simple explanation, a method of providing a legal document clustering search service using artificial intelligence has been illustrated and described as a series of blocks, but the present invention is not limited to the order of the blocks, and some blocks are different from those shown and operated in this specification. It may be performed in sequence or may be performed simultaneously. In addition, new blocks not described in the present specification and drawings may be added, or some blocks may be deleted or changed.

As mentioned above, although embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art to which the present invention pertains can realize that the present invention can be embodied in other specific forms without changing its technical spirit or essential features. you will be able to understand Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive.

100: Legal document clustering search service providing server (server)
200: user terminal
300 : external server
400: network

Claims (12)

A method performed by a computing device, comprising:
receiving one or more search terms;
generating one or more dynamic clusters by performing dynamic clustering on a plurality of pre-stored legal documents using the input one or more search words;
selecting one or more static clusters related to the one or more input keywords from among a plurality of static clusters generated by static clustering the plurality of pre-stored legal documents; and
providing search results comprising the generated one or more dynamic clusters and the selected one or more static clusters;
A method of providing legal document clustering search service using artificial intelligence. According to claim 1,
The step of creating the one or more dynamic clusters comprises:
selecting a plurality of legal documents including the input at least one search word from among the plurality of pre-stored legal documents;
vectorizing the selected plurality of legal documents; and
Calculating a degree of similarity between the plurality of vectorized legal documents, and performing dynamic clustering on the selected plurality of legal documents based on the calculated degree of similarity;
A method of providing legal document clustering search service using artificial intelligence. 3. The method of claim 2,
The step of selecting a plurality of legal documents including the input one or more search terms,
Selecting a legal document containing the one or more inputted search words at least once from among the plurality of pre-stored legal documents, and when the number of the selected legal documents exceeds a preset maximum value, the number of selected legal documents increases Including the step of sequentially increasing the minimum number of inclusions of the one or more inputted search terms for the plurality of pre-stored legal documents until it is less than or equal to the preset maximum value,
A method of providing legal document clustering search service using artificial intelligence. According to claim 1,
The step of creating the one or more dynamic clusters comprises:
When a plurality of search terms are input, a legal document including all of the plurality of search terms is selected from among the plurality of pre-stored legal documents, and dynamic clustering is performed on the selected legal documents,
When the selected legal document is less than a preset minimum value, a priority among the plurality of search terms until the number of the selected legal documents is equal to or greater than the preset minimum value - The priority is the order in which the plurality of search words are input and the plurality of It is set based on at least one of the operators input with the search word of - including sequentially excluding from the search word with the lowest
A method of providing legal document clustering search service using artificial intelligence. According to claim 1,
The method further comprises generating a plurality of static clusters by performing static clustering on the plurality of pre-stored legal documents,
The step of creating the plurality of static clusters includes:
By performing static clustering according to each of the plurality of pre-stored legal documents as a reference document, the reference document includes at least one of precedents, statutes, documents, and comments referred to by each of the plurality of pre-stored legal documents. creating a first static cluster of
By performing static clustering on each of the plurality of pre-stored legal documents according to a related law, the related law includes at least one of laws and articles cited by each of the plurality of pre-stored legal documents, by performing static clustering, a plurality of second static creating a cluster; and
generating a plurality of third static clusters by performing static clustering of each of the plurality of pre-stored legal documents according to an issue dealt with by each of the plurality of pre-stored legal documents,
A method of providing legal document clustering search service using artificial intelligence. According to claim 1,
The step of selecting one or more static clusters comprises:
calculating a relationship between the plurality of static clusters and the inputted one or more search terms; and
selecting one or more static clusters in which the calculated relevance is equal to or greater than a preset value;
A method of providing legal document clustering search service using artificial intelligence. 7. The method of claim 6,
The step of selecting one or more static clusters in which the calculated similarity is equal to or greater than a preset value includes:
When a plurality of search words are input, one or more static clusters having relevance to all search words from among the plurality of static clusters are selected, or an average value of relevance for each of the plurality of search words is greater than or equal to the preset value. Selecting more than one static cluster,
A method of providing legal document clustering search service using artificial intelligence. According to claim 1,
Providing the search results comprises:
When the number of dynamic clusters generated using the first search word and static clusters selected using the first search word is less than a preset number, a second search word that is a synonym of the first search word or a third keyword that is a similar word to the first search word A dynamic cluster generated using a search term and a static cluster selected using the second search term or the third search term are provided together.
A method of providing legal document clustering search service using artificial intelligence. According to claim 1,
The step of receiving the input of the one or more search terms includes:
providing a plurality of query contents to the user in response to receiving a request from the user to provide a legal document search service;
receiving a plurality of answers to the provided plurality of questions from the user; and
extracting one or more keywords by analyzing the plurality of input answers, and selecting the extracted one or more keywords as a search word;
A method of providing legal document clustering search service using artificial intelligence. According to claim 1,
Collecting a plurality of legal documents, further comprising the step of processing and storing the collected plurality of legal documents,
The step of processing and storing the plurality of collected legal documents is,
When the plurality of collected legal documents are legal documents in the form of images, analyzing the legal documents in the form of images to extract one or more character strings, processing and storing the extracted one or more character strings on a preset template ; and
When the language of the plurality of collected legal documents is a foreign language, automatically translating the foreign language into a preset language, and processing and storing the automatically translated legal document on the preset template;
A method of providing legal document clustering search service using artificial intelligence. processor;
network interface;
Memory; and
a computer program loaded into the memory and executed by the processor;
The computer program is
an instruction for receiving one or more search terms;
instructions for generating one or more dynamic clusters by performing dynamic clustering on a plurality of pre-stored legal documents using the input one or more search words;
an instruction for selecting one or more static clusters related to the one or more input keywords from among a plurality of static clusters generated by static clustering the plurality of pre-stored legal documents; and
instructions for providing search results including the generated one or more dynamic clusters and the selected one or more static clusters;
Legal document clustering search service providing server using artificial intelligence. combined with a computing device,
receiving one or more search terms;
generating one or more dynamic clusters by performing dynamic clustering on a plurality of pre-stored legal documents using the input one or more search words;
selecting one or more static clusters related to the one or more input keywords from among a plurality of static clusters generated by static clustering the plurality of pre-stored legal documents; and
stored in a computer-readable recording medium for executing the step of providing a search result including the generated one or more dynamic clusters and the selected one or more static clusters;
A computer program recorded on a computer-readable recording medium.

Images