KR102566991B1 - System for providing countersign service using fingerprint - Google Patents

Abstract

An electronic service providing system using a fingerprint is provided, at least one user terminal for uploading a contract and requesting an electronic seal, a reception unit for receiving a contract from at least one user terminal, and a finger pre-built on each page constituting the contract A setting unit that sets to generate a fingerprint with a print generation parameter, a content storage unit that creates a fingerprint on the contents of a page and stores it as a content fingerprint, and generates a fingerprint on the order of the page number of the page and an electronic service providing server including a sequence storage unit for storing sequence fingerprints, and a page storage unit for writing content fingerprints and sequence fingerprints as page fingerprints.

Description

Electronic service providing system using fingerprint {SYSTEM FOR PROVIDING COUNTERSIGN SERVICE USING FINGERPRINT}

The present invention relates to a system for providing electronic interpersonal services using fingerprints, which generates fingerprints for the content and order of each page, and guarantees continuity by following and following each page, thereby preventing forgery and alteration. to provide.

An electronic signature refers to an electronic signature or seal applied to an electronic document in the same way as signing or stamping a paper document. E-commerce, e-financial transactions, and e-government services can be used non-face-to-face by using e-signatures. Until now, the most widely used electronic signature method by the general public has been a public certificate. A public certificate is a technically superior electronic signature means, but it was also an inconvenient electronic signature means for subscribers and users. The overly concentrated legal system design and technology focus on certified electronic signatures were recognized as obstacles to various innovative services, and in 2014, the Electronic Financial Transaction Act, which deletes the regulations used as the basis for forcing the use of accredited certificates in electronic financial transactions From the revision to the full revision of the Electronic Signature Act in 2020, a series of accredited electronic signature regulations, such as accredited certificates, accredited certification services, and accredited certification authorities, have been completely deleted. Public opinion wanted convenience rather than safety, and a new electronic signature system was implemented through law revision, and various private certificates are appearing in the market with innovative services in accordance with the purpose of the revised electronic signature system. Various services such as pass certificates from the three mobile telecommunication companies, Kakao Pay certificates from Kakao, bank signs affiliated with banks, Toss certificates from simple remittance service Toss, and Naver certificates from Naver are being introduced to the market.

At this time, a method of providing a function of electronic signature by performing batch packaging of electronic signatures or printing a location for stamping has been researched and developed. In this regard, Korea Patent Registration No. 10-0886690 ( Announced on March 4, 2009) and Korean Registered Patent No. 10-0563076 (Announced on March 24, 2006), when conducting an electronic contract, package the contract file to be compressed into a single file, and batch electronic signature a configuration that provides a function of sealer to ensure that some of the documents are not missing, converts multiple documents into electronic documents, and places an electronic seal on the back and front sides of the front and back sheets where the front and back sides of the document are folded; A configuration for setting coordinates and storing the set position and direction of an electronic interpersonal person in a database is disclosed.

However, in the case of the former, the contract file is compressed, but it is not a configuration that provides an electronic seal to ensure the continuity and integrity of the contract. It is not a composition that proves the continuity of the document.

Unlike traditional signatures or contracts using paper, signatures or electronic contracts do not require an intersign or seal because all procedures are completed with one electronic original document and legal effect takes effect.

In addition, since anti-counterfeiting and security technologies such as audit trails, which can only be provided by electronic missions, are basically applied to Hancom Sign, you can sign or contract safely. The seal of paper signature is a method of sealing used to prove the continuity of each page when a contract consists of several pages. tell me how to do it In addition, the paper signature seal is a method of sealing that is used to prove the interrelationship of the document when there are multiple parties to the contract. The purpose of the seal and seal is to prevent forgery and falsification, but also to prevent the risk of missing or adding one sheet when there are multiple contracts and to ensure continuity. Therefore, research and development of electronic seals capable of ensuring the continuity of documents and preventing forgery and alteration are required.

In one embodiment of the present invention, if there is no fingerprint of the previous page using a fingerprint generation parameter for each page from the first page to the last page of a document consisting of a plurality of pages, it is filled with a null value, After generating the fingerprint of the body of the current page, if there is no fingerprint of the next page, fill it with a null value. When the content of one page is confirmed, the content fingerprint is created, and the order of the pages Generates a fingerprint, and after generating a page fingerprint with a content fingerprint and a sequence fingerprint for any one page, move to the next page and repeat the process to ensure the continuity of the document, and to ensure the continuity of the document Electronic signature using fingerprint, providing electronic authentication to prevent addition or deletion of documents by setting the previous page to point to the next page accordingly, and setting each page to follow and follow A service provision system can be provided. However, the technical problem to be achieved by the present embodiment is not limited to the technical problem as described above, and other technical problems may exist.

As a technical means for achieving the above-described technical problem, an embodiment of the present invention, at least one user terminal for uploading a contract and requesting an electronic signature, and a receiving unit for receiving the contract from at least one user terminal, constituting the contract A setting unit that sets a fingerprint to be generated with a fingerprint generation parameter pre-established in each page, a content storage unit that creates and stores a fingerprint about the content of a page as a content fingerprint, and a sequence that is the page number of the page and an electronic interpersonal service providing server including a sequence storage unit that generates and stores a fingerprint related to a sequence fingerprint, and a page storage unit that writes a content fingerprint and a sequence fingerprint as a page fingerprint.

According to any one of the above-described problem solving means of the present invention, if there is no fingerprint of the previous page by using a fingerprint generation parameter for each page from the first page to the last page of a document consisting of a plurality of pages, null ( Null), create a fingerprint of the body of the current page, and then fill it with a null value if there is no fingerprint of the next page. Generates a sequence fingerprint for the sequence of pages, and after generating a page fingerprint with a content fingerprint and sequence fingerprint for an entire page, move to the next page and repeat the process to ensure the continuity of the document , In accordance with the order of each page, the previous page is set to point to the next page, and by setting each page to Follow and Follow, electronic signatures can be provided to prevent the addition or deletion of documents. there is.

1 is a diagram for explaining an electronic service providing system using a fingerprint according to an embodiment of the present invention.
FIG. 2 is a block diagram illustrating an electronic interpersonal service providing server included in the system of FIG. 1 .
3 and 4 are diagrams for explaining an embodiment in which an electronic interpersonal service using a fingerprint according to an embodiment of the present invention is implemented.
5 is an operational flowchart illustrating a method for providing electronic interpersonal services using fingerprints according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail so that those skilled in the art can easily practice the present invention with reference to the accompanying drawings. However, the present invention may be embodied in many different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

Throughout the specification, when a part is said to be "connected" to another part, this includes not only the case where it is "directly connected" but also the case where it is "electrically connected" with another element interposed therebetween. . In addition, when a part "includes" a certain component, this means that it may further include other components, not excluding other components, unless otherwise stated, and one or more other characteristics. However, it should be understood that it does not preclude the possibility of existence or addition of numbers, steps, operations, components, parts, or combinations thereof.

As used throughout the specification, the terms "about", "substantially", etc., are used at or approximating that value when manufacturing and material tolerances inherent in the stated meaning are given, and do not convey an understanding of the present invention. Accurate or absolute figures are used to help prevent exploitation by unscrupulous infringers of the disclosed disclosure. The term "step of (doing)" or "step of" as used throughout the specification of the present invention does not mean "step for".

In this specification, a "unit" includes a unit realized by hardware, a unit realized by software, and a unit realized using both. Further, one unit may be realized using two or more hardware, and two or more units may be realized by one hardware. On the other hand, '~ unit' is not limited to software or hardware, and '~ unit' may be configured to be in an addressable storage medium or configured to reproduce one or more processors. Thus, as an example, '~unit' refers to components such as software components, object-oriented software components, class components, and task components, processes, functions, properties, and procedures. , subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays and variables. Functions provided within components and '~units' may be combined into smaller numbers of components and '~units' or further separated into additional components and '~units'. In addition, components and '~units' may be implemented to play one or more CPUs in a device or a secure multimedia card.

In this specification, some of the operations or functions described as being performed by a terminal, device, or device may be performed instead by a server connected to the terminal, device, or device. Likewise, some of the operations or functions described as being performed by the server may also be performed by a terminal, apparatus, or device connected to the server.

In this specification, some of the operations or functions described as mapping or matching with the terminal mean mapping or matching the terminal's unique number or personal identification information, which is the terminal's identifying data. can be interpreted as

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

1 is a diagram for explaining an electronic service providing system using a fingerprint according to an embodiment of the present invention. Referring to FIG. 1 , an electronic service providing system 1 using a fingerprint may include at least one user terminal 100, an electronic service providing server 300, and at least one contractor terminal 400 there is. However, since the electronic service providing system 1 using the fingerprint of FIG. 1 is only one embodiment of the present invention, the present invention is not limitedly interpreted through FIG. 1 .

At this time, each component of FIG. 1 is generally connected through a network (Network, 200). For example, as shown in FIG. 1 , at least one user terminal 100 may be connected to an electronic service providing server 300 through a network 200 . In addition, the electronic service providing server 300 may be connected to at least one user terminal 100 and at least one contractor terminal 400 through the network 200 . In addition, at least one contractor terminal 400 may be connected to the electronic service providing server 300 through the network 200 .

Here, the network refers to a connection structure capable of exchanging information between nodes such as a plurality of terminals and servers, and examples of such networks include a local area network (LAN) and a wide area network (WAN: Wide Area Network), the Internet (WWW: World Wide Web), wired and wireless data communications networks, telephone networks, and wired and wireless television communications networks. Examples of wireless data communication networks include 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 , Internet (Internet), LAN (Local Area Network), Wireless LAN (Wireless Local Area Network), WAN (Wide Area Network), PAN (Personal Area Network), RF (Radio Frequency), Bluetooth (Bluetooth) network, NFC ( A Near-Field Communication (Near-Field Communication) network, a satellite broadcasting network, an analog broadcasting network, a Digital Multimedia Broadcasting (DMB) network, etc. are included, but not limited thereto.

In the following, the term at least one is defined as a term including singular and plural, and even if at least one term does not exist, each component may exist in singular or plural, and may mean singular or plural. It will be self-evident. In addition, the singular or plural number of each component may be changed according to embodiments.

At least one user terminal 100 may be a terminal of a user who uploads a contract using a web page, app page, program or application related to an electronic seal service using a fingerprint and requests electronic seal.

Here, at least one user terminal 100 may be implemented as a computer capable of accessing a remote server or terminal through a network. Here, the computer may include, for example, a laptop, a desktop, a laptop, and the like equipped with a navigation system and a web browser. In this case, at least one user terminal 100 may be implemented as a terminal capable of accessing a remote server or terminal through a network. At least one user terminal 100 is, for example, a wireless communication device that ensures portability and mobility, navigation, PCS (Personal Communication 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 ) may include all types of handheld-based wireless communication devices such as terminals, smartphones, smart pads, tablet PCs, and the like.

The electronic personal service providing server 300 may be a server that provides an electronic personal service web page, app page, program, or application using a fingerprint. Also, the electronic service providing server 300 may be a server that receives a contract between the user terminal 100 and the contractor terminal 400 and provides electronic service. To this end, the electronic service providing server 300 creates a fingerprint for the contents and order of a plurality of pages constituting the contract, and establishes a following and following relationship so that the previous and next pages are connected, thereby ensuring original integrity. It may be a server that maintains and proves the continuity of pages so that pages are not added or omitted.

Here, the electronic service providing server 300 may be implemented as a computer capable of accessing a remote server or terminal through a network. Here, the computer may include, for example, a laptop, a desktop, a laptop, and the like equipped with a navigation system and a web browser.

At least one contracting party terminal 400 may be a terminal of a user's contracting party using a web page, app page, program or application related to an electronic service using a fingerprint.

Here, at least one contractor terminal 400 may be implemented as a computer capable of accessing a remote server or terminal through a network. Here, the computer may include, for example, a laptop, a desktop, a laptop, and the like equipped with a navigation system and a web browser. In this case, at least one contractor terminal 400 may be implemented as a terminal capable of accessing a remote server or terminal through a network. At least one contractor terminal 400 is, for example, a wireless communication device that ensures portability and mobility, navigation, PCS (Personal Communication 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 ) may include all types of handheld-based wireless communication devices such as terminals, smartphones, smart pads, tablet PCs, and the like.

2 is a block configuration diagram for explaining an electronic service providing server included in the system of FIG. 1, and FIGS. 3 and 4 are an implementation of an electronic service using a fingerprint according to an embodiment of the present invention. It is a drawing for explaining an example.

Referring to FIG. 2, the electronic service providing server 300 includes a reception unit 310, a setting unit 320, a content storage unit 330, an order storage unit 340, a page storage unit 350, and a creation management unit. 360, relationship connection unit 370, pointer unit 380, repetition unit 390, authentication unit 391, content hashing unit 393, sequence hashing unit 395, legal tech unit 397 and an automatic writing unit 399.

The electronic interpersonal service providing server 300 according to an embodiment of the present invention or another server (not shown) operating in conjunction with at least one user terminal 100 and at least one contractor terminal 400 using a fingerprint When transmitting an electronic service application, program, app page, web page, etc., at least one user terminal 100 and at least one contractor terminal 400 are electronic service applications, programs, and app pages using a fingerprint , install or open web pages, etc. In addition, the service program may be driven in at least one user terminal 100 and at least one contractor terminal 400 using a script executed in a web browser. Here, the web browser is a program that allows users to use the web (WWW: World Wide Web) service, and means a program that receives and displays hypertext described in HTML (Hyper Text Mark-up Language). For example, Netscape , Explorer, Chrome, and the like. In addition, an application means an application on a terminal, and includes, for example, an app running on a mobile terminal (smart phone).

Referring to FIG. 2 , the acceptance unit 310 may receive a contract uploaded from at least one user terminal 100 . At least one user terminal 100 may upload a contract and request an electronic seal. At this time, it may be uploaded in the form of a file or a scanned copy of a paper document created offline may be uploaded.

The setting unit 320 may set to generate a fingerprint using a fingerprint generation parameter pre-established in each page constituting the contract. The dictionary meaning of fingerprint is a fingerprint, and it usually means a technology for identifying illegally distributed digital content. It is also referred to as an encryption algorithm and a hash algorithm, which means an encryption key or hash value, which is called an encryption or message fingerprint. At this time, the fingerprint is used in a way similar to a watermark to be inserted to prevent unauthorized copying of digital data. It is a technology that protects buyers from illegally distributing digital data by providing a post-detection function that enables Fingerprint is similar in that a binary mark is inserted into data to prevent illegal copying of digital data, but since the content of the mark inserted for data is all different, an additional function to identify each of the same data is provided. . The Fingerprint Algorithm is an algorithm that maps arbitrary files or large amounts of data to shorter bits, Fingerprints. Fingerprints are used to quickly identify the original and transmit or compare large amounts of data. you can avoid doing For example, when trying to check whether a remote large-capacity file has been changed in a web browser, the change comparison can be performed efficiently by receiving the fingerprint of the file and comparing it with the previously stored fingerprint. The Fingerprint Function is a high-speed hash function used in the field of verifying data in applications that do not require advanced encryption such as the Cryptographic Hash Function.

The content storage unit 330 may generate and store a fingerprint related to page content as a content fingerprint. The order storage unit 340 may generate and store a fingerprint related to the order, which is a page number of a page, as an order fingerprint. Recent sound source search technology uses a method of extracting a fingerprint using feature points extracted through frequency analysis. In the case of an image, low-frequency components include structure information and high-frequency components include texture information, so frequency analysis It is useful for extracting meaningful information from images as well as from music. When fingerprints identify copyrighted works, feature points are extracted by converting image data into a frequency domain. The extracted feature points are grouped into consecutive groups, and an address value is generated using the positional relationship between the representative feature point of the group and the element feature point. The generated address value becomes a key representing a specific picture area of the image. For work identification, the database is searched for the address value generated from the piece of work, and the work in which the positional relationship is sequentially matched is selected as the original work. In this way, a piece of a work that is part of the original work can be identified as a work. Using this, in an embodiment of the present invention, a contract fingerprint may be generated using a frequency domain feature.

<Create fingerprint>

Linear interpolation can be used in the process of normalizing the image size for frequency feature point extraction. The linear interpolation algorithm may determine a corresponding pixel value by referring to neighboring pixel values. When an outer pixel is lost or a framed or reduced image is input, a high frequency component is changed. In the case of contracts, since the size of the original image does not vary greatly, the degree of high-frequency distortion due to image size normalization is relatively small. In the case of extracting a frequency feature point at a fixed position in all frequency domains, there is a disadvantage in that it is very vulnerable to distortion such as outer pixel loss. In addition, since the uniqueness of positional relationship information between feature points is weak, identification speed and accuracy decrease as the number of contracts included in the database increases. Accordingly, in one embodiment of the present invention, a low-frequency contract fingerprint generation method having improved identification accuracy and speed compared to general fingerprints can be used.

<Extract ROI>

In the ROI (Region of Interest) extraction step, the picture area of the image is extracted. In the image normalization step, the vertical length of the extracted ROI is normalized in order to extract fingerprints of the same dimension from contract images that have been modified and posted in various sizes. In the spectrogram generation step, the contract image is transformed into the frequency domain in units of rows. The step of generating a fingerprint is a step of extracting a fingerprint from a low frequency region of the spectrogram. The first step in contract fingerprint generation is the process of extracting ROI by dividing a given contract piece into frames. In general identification algorithms, feature points are extracted by simply normalizing the size of the input contract fragments, but if the entire contract fragment is used as an input, not only unnecessary data is generated due to blanks, but also the probability of misclassification due to blanks increases. A method of removing blanks and extracting only contract frames that actually contain content can be used. Therefore, as a preprocessing process, only the drawing area excluding the blank space is extracted from the image. For ROI extraction, the blanks are removed using the second partial derivative of all rows of the contract image. If the top and bottom blanks are removed through row-by-row operation, the left and right blanks are removed by obtaining the second partial derivative of all columns. When obtaining the 2nd partial derivative of the contract image, the 2nd partial derivative value is very small because the change in pixel value is small or constant for blanks with a single color or constant gradation. When the horizontal length of the image is W, the vertical length is H, the blanking threshold in row units is εRow, and the blanking threshold in column units is εcloumn, in consideration of the noise present in the image, 2 of one row or column as in Equation 1 If the sum of the differential partial derivative values is below a very small certain threshold value, it is regarded as a margin.

ROI is extracted frame by frame from the contract image entered through this process.

<Image normalization and spectrogram extraction>

A general algorithm extracts frequency feature points by matching the horizontal length of an image, but the method according to an embodiment of the present invention generates one ROI as one row-dimensional vector. In order to extract fingerprints of the same dimension from ROIs of various sizes, the vertical length of the ROI is normalized to Hnormpixel. In order to maintain the image ratio, the normalized horizontal length Wnorm is calculated as in Equation 2.

As in the general method, the size-normalized ROI is transformed into a frequency domain by performing a 1-dimensional Fourier transform (1D-FFT) for each row.

<Generation of low-frequency fingerprints>

Since low-frequency components contain most of the image information, compression efficiency is improved by using a method of removing high-frequency components even in the JPEG compression method. Therefore, it is appropriate to use a low-frequency component as a strong feature representing an image. For example, when analyzing the low-frequency Fourier transform value according to the ROI row of the contract image, the high-frequency domain Fourier transform value changes greatly with distortion, whereas the low-frequency domain Fourier transform value does not change significantly with distortion.

In one embodiment of the present invention, it is possible to introduce a fingerprint extraction algorithm based on similarity between vectors by generating one vector having ROI rows in each dimension. In the case of simply setting the maximum value of the Fourier transform in the entire frequency domain as a fingerprint, it is difficult to secure the uniqueness of the fingerprint representing the contract because most DC (Direct Current) component values are extracted. To maximize uniqueness, select the maximum value of the Fourier transform of each row in the ROI-specific low-frequency region (α*width to β*width). As a result, one Hnorm-dimensional fingerprint vector can be generated in one ROI. For contract identification, a vector similarity between a plurality of fingerprints stored in a database and a fingerprint extracted from an input query image is calculated. For example, a Euclidean distance may be used for similarity calculation. Of course, it is not limited thereto.

The page storage unit 350 may write a content fingerprint and a sequence fingerprint as a page fingerprint. In the page storage unit 350, the contents and order are created as fingerprints and then stored, and then moved to the next page and repeated until the last page appears.

The creation management unit 360 fills the sequence fingerprint with a null value if there is no previous page and the sequence fingerprint with a null value if there is no next page based on any one page of the plurality of pages constituting the contract. there is. For example, assuming that a contract exists from page 1 to page N, if it is page 1, there is no previous page, so the previous sequence fingerprint of page 1 is a null value.

The relationship linking unit 370, based on any one page of the plurality of pages constituting the contract, sets the next page of any one page to follow any one page, and transfers any one page. You can set the page to follow. The pointer unit 380, based on any one page of the plurality of pages constituting the contract, sets the next page of any one page to point to any one page, and any one page points to any one page. can be set to point to the next page of If you set the back-and-forth relationship in this way, you can quickly know when a page is missing, modified, or added.

The repetition unit 390 may repeat the page fingerprint generation process until the last page of the contract.

When reaching the last page of the contract, the authentication unit 391 may authenticate that the contract is the last page and end the creation process.

The content hashing unit 393 may convert page content including text and images of each page into a hash function and store a hash value. The sequence hashing unit 395 may store the hash value by converting the number content of the sequence of pages into a hash function. As described above, a fingerprint may be included as a watermark, but a hash value may be generated by converting the content of the page itself into a hash function. Since this may vary depending on the embodiment, it is not limited to any one. In addition, it will be apparent that various technologies other than the listed technologies may be used.

<hash value>

A hash value can be generated by converting an image corresponding to the content into a hash function. At this time, the hash algorithm is a method for constructing a hash function, and the hash algorithm is a deterministic algorithm, and the result value is also different if the original data is slightly different. However, if the original message is known, generating the ciphertext is a quick and easy process. The essential elements of a hash function are antiphase resistance, second antiphase resistance, and collision resistance. Through this, the hash can be used to make it difficult to find the original with only hash or an attack that modifies the input value, or to find the same input value for different input values. guaranteed However, differences between each element may occur depending on the form of hash algorithm. The most effective way to verify the integrity of the content is to use a cryptographically secure one-way hash function, and the hash function is widely used in the information protection field, such as digital signatures and message authentication codes. A process of verifying integrity using a hash function may be as shown in Table 1 below.

detail →Send→ detail ↓ ↓ hash function hash function ↓ ↓ hash code hash code └→ comparison ←┘

<Hash Function Repetition> The repetition principle of hash function is also called MD (Merkle-Damgard) or domain extension or operating mode. This is a method of constructing a hash function that handles arbitrary inputs using a function with a fixed input/output size. Here, a function having a fixed input/output size is called a compression function. If the compression function has collision resistance, the hash function configured in the MD method can have the same collision resistance as long as appropriate padding is used. Here, the message is decomposed into short message blocks, then one at a time. The designed compression function is serially connected to sequentially process message blocks. In the case of padding, it is called MD-Strengthening in the form of adding a certain padding block to a given message. If this is applied in the case of SHA-1, generally, 1 must be appended to the end of the message and then 0 is concatenated so that the entire block is the compression function. It is processed so that it is a multiple of the message block length. Here, the remaining 64 bits are filled with the length of the message, and the length of the message is added just before output. Through this same principle, the hash can have a fixed output value and prevent the same output value from appearing in different input values. <SHA>

SHA (Secure Hash Algorithm) was used in many security protocols and programs, such as TLS, SSL, PGP, SSH, and IPSec, but was replaced with SHA-256 after a vulnerability was discovered. SHA-256, the current standard, has a similar structure to other SHA-2 families, but performs 64 rounds of hash operation, and outputs a fixed value of 256 bits as a result of the operation, which is called a message digest. SHA-1 and SHA-2 use a similar method with only differences in blocks or repetitions, but SHA-256 updates 8 32 bits based on compression and has a corresponding state variable, updating a total of 64 times. It completes the message digest in an iterative fashion. Two linear functions (Σ), fmaj and fIF, which are bit Boolean functions for step conversion, appear, and through them, values are changed and compression is performed. Here, ROTR is to bit-rotate x to the right by n positions.

Overall, the ith step uses fixed constants Ki and Wi, and Wi can take the value of a previously generated message. Through this process, the same message digest can be output regardless of the input value of any length. SHA outputs a different hash value if even one letter is incorrect, so it can be useful for verifying document integrity. Although the above method has been described based on content, a hash value for a text corresponding to a page number may be generated in the following order.

The legal tech department 397 may review the risk after reviewing the contents of the contract with a pre-built artificial intelligence algorithm before including the electronic person after the contract is uploaded. When contracts between private parties or between companies are carried out, transactions are conducted based on the provisions specified in the contract. However, if a poison clause that can be applied to one's disadvantage is not discovered before signing the contract or even if it is recognized, if it is overlooked and not dealt with, an issue arises during the transaction execution stage, causing problems such as work delay or worsening profit and loss. . In order to prevent such a problem, the person in charge of the contract or related experts reviews the contract before signing the contract to find out the toxic clause and take appropriate measures, but this consumes a lot of time and money. Recently, research on an automated system that can automatically determine poisonous clauses in a contract is active, but since it relies on a rule-based method, there are many improvements compared to direct review by experts, but a large amount of time is spent on developing the system. Circumstances and costs do not change.

Therefore, in one embodiment of the present invention, a deep learning-based text classification model capable of determining poisonous clauses by reviewing the contract in advance can be used to minimize the above-mentioned time and cost problems. Text classification, a problem that has been frequently dealt with in the field of natural language processing, is a task to classify a given sentence, paragraph or document to which class it belongs to, and is used for sentiment analysis, news article classification, and spam mail classification. In text classification work, when an input unit consists of several sentences (paragraphs, documents, etc.), sentences that are not related to the core content sometimes act as noise for accurate classification. In this way, a process of selecting sentences that can act as noise in classification is necessary, so a paragraph consisting of a large number of sentences is received as an input, the weight of each sentence is adjusted and reflected, and based on this, whether or not there is a poison clause is determined. You can use a deep learning-based text classification model that classifies text.

<Risk Detection>

In the encoding step, a sentence encoder (Sentence Encoder) may generate a sentence vector corresponding to each sentence in a paragraph. The class embedding layer receives each sentence vector as an input and outputs similarity information between each sentence vector and the class to be classified. The gate layer receives a concatenated value of each sentence vector and similarity information with a class corresponding to the sentence vector, calculates a weight for each sentence, and outputs a sentence vector reflecting the weight. The weighted sentence vector is used as an input to the Paragraph Encoder to generate one paragraph vector for multiple sentence vector inputs. In the decoding step, the generated short vector is used as an input to the class embedding layer, and similarity information between the corresponding short vector and classes to be classified is output. Finally, the forward propagation neural network receives a value obtained by connecting the paragraph vector and the output value of the class embedding layer, and generates a probability distribution about which class the input paragraph belongs to.

As the sentence encoder, for example, an ALBERT-based sentence encoder may be used. The output value of the encoder corresponding to the [SEP] token for distinguishing each sentence can be used as a sentence vector for each sentence. The sentence encoder receives each sentence in a paragraph as an input and generates a sentence vector corresponding to each sentence. The class embedding layer receives sentence vectors or paragraph vectors and projects them into the class vector space through the forward propagation neural network. At this time, the projected vector is called an embedding vector, and the embedding vector and the class embedding matrix generate a cosine similarity distribution of each class vector to be classified and the embedding vector through a matrix multiplication operation. Finally, the class embedding layer outputs a value obtained by connecting the embedding vector and the cosine similarity distribution.

The gate layer is a method that applies the gate theory of recurrent neural networks, and adjusts the reflection ratio of each sentence by using a sentence vector, a corresponding embedding vector, and a value obtained by connecting the embedding vector and the cosine similarity distribution of each class vector. do. Through this process, the sentences in the input paragraph have different weights for each sentence based on the semantic information of the sentence, the location information on the class vector space, and the association information with each class. A paragraph encoder generates paragraph vectors for all input paragraphs by using each weighted sentence vector as an input to a single gated recurrent unit (GRU) layer to which the attention-focusing method is applied. After building the model in this way, when creating a dataset, a dataset can be created by pairwise mapping the paragraphs extracted from the actual contract and whether or not there are poison clauses corresponding to them, and modeling the artificial intelligence algorithm by using it for learning and verification. can do.

When creating a contract in the user terminal 100, the automatic creation unit 399 uploads the contents of the contract including the purpose, use, and contractor information from the user terminal 100, recommends the type of contract, and recommends the type of contract. ), if you select the type of contract, you can automatically create a contract with the contents of the contract. Recently, Open AI, which generates sentences using GPT-3, has been announced. When using this, users can freely create sentences they want. Based on this, if the user uploads information about what purpose and purpose the contract is to be drawn up and contractor information, the contract can be created automatically.

In addition, an embodiment of the present invention may further use a method of determining the degree of similarity of documents in order to confirm the originality of the document with respect to the contract attached by the electronic person as a fingerprint. For example, Locality Sensitive Hashing (LSH) may be used. In general, similarity is a scale that can be distinguished as [similar] or [not similar] with the naked eye. At this time, the judgment of [similar] and [not similar] is a subjective indicator that can cause differences from person to person. If such a subjective index determines whether a contract is original or not, the result is inevitably subjective. However, since the above-described LSH is programmed, implemented, and processed through a mathematically established formula, it can be free from misunderstanding or stigma of subjective similarity. LSH is one of the similarity comparison algorithms and is used in fields such as data mining, pattern recognition, computer vision, computational geometry, and data compression. It is also used to check spelling, detect plagiarism and measure chemical similarities.

<Shingling>

The Shingling step means converting a document into a set of words. It is the same process as representing each document as a number in a vector similarity model. A document is expressed as a kind of word (character) sequence divided according to the number of tokens defined as characters, words, or other units. In order to compress the document expressed in this way, the string of the document is hashed with a specified byte, and the result of this entire hashing is expressed as a document. If hashing is done in word or character units, intuitively similar documents have the same hashing result in common when viewed as a whole document.

<Min-Hashing>

The Min-Hashing step is a step through the MinHash algorithm. This step is a process of converting a large document set (Shingles processed documents) into a short signature while maintaining the similarity of documents. In the process of Shingles, the document is expressed as a set of Shingles, and this set is expressed like a matrix (column, row) and converted into a new signature array using permutation π. The definition of MinHash used at this time is as follows: Same as Equation 4.

In the above formula, for permutation π, the value of the first column is a value expressing the number of rows in the first (order of permutation). By applying this to all columns, several randomly selected permutations π are applied to create signatures for each column. The result is expressed as a matrix, and for permutation π, columns are Sets and rows are expressed as MinHash values. Since the signatrue generated in this way has a fixed value regardless of the size of the original document, it is possible to compare the degree of similarity with documents of different sizes.

<LSH>

The whole idea of Locality Sensitive Hashing (LSH) starts with hash columns for matrices generated by the MinHash process. Even after going through the MinHash process, it can be seen that the hash column appears as many times as the number of documents. LSH is a concept that compares equality by roughly calculating only columns with similar hash results and putting them in the same buckets in order not to compare all documents. Assuming that it is defined as a band consisting of each column of matrix M, each band calculates a part of each column as a hash table with buckets. Bands reaching the same area become document pairs with a similarity probability, and since only document pairs with such a probability are compared, the number of comparisons of all documents can be reduced to compare the similarity between large amounts of electronic information.

Hereinafter, an operation process according to the configuration of the above-described inter-electronic service providing server of FIG. 2 will be described in detail with reference to FIGS. 3 and 4 as examples. However, it will be apparent that the embodiment is only any one of various embodiments of the present invention, and is not limited thereto.

Referring to FIG. 3, (a) the electronic service providing server 300 receives a contract from the user terminal 100 or the contractor terminal 400, and prints a fingerprint for the contents and order of each page as shown in (b). After creating, create a structure that connects the front and back of each page as shown in (c). (d) The electronic intersign service providing server 300 transmits the electronic intersigned contract to the user terminal 100 and the contractor terminal 400. When this is schematized, it is shown in FIG. 4 .

Matters that have not been described for the method of providing electronic interpersonal services using fingerprints of FIGS. 2 to 4 are the same as those described for the method of providing interpersonal services using fingerprints through FIG. 1, or from the described content. Since it can be easily inferred, the following description will be omitted.

FIG. 5 is a diagram illustrating a process of transmitting/receiving data between components included in the electronic service providing system using the fingerprint of FIG. 1 according to an embodiment of the present invention. Hereinafter, an example of a process of transmitting and receiving data between each component will be described through FIG. 5, but the present application is not limited to such an embodiment, and according to various embodiments described above, It is obvious to those skilled in the art that a process of transmitting and receiving data may be changed.

Referring to FIG. 5 , the electronic service providing server receives a contract uploaded from at least one user terminal (S5100).

Then, the electronic service providing server is configured to generate a fingerprint using a fingerprint generation parameter pre-established in each page constituting the contract (S5200).

In addition, the electronic service providing server generates a fingerprint related to the content of the page and stores it as a content fingerprint (S5300).

Then, the electronic service providing server generates a fingerprint related to the order, which is the page number of the page, and stores it as the order fingerprint (S5400), and writes the content fingerprint and the order fingerprint as the page fingerprint (S5500).

The order between the above-described steps (S5100 to S5500) is only an example, and is not limited thereto. That is, the order of the above-described steps (S5100 to S5500) may be mutually changed, and some of the steps may be simultaneously executed or deleted.

Matters that have not been explained about the method of providing electronic interpersonal services using the fingerprint of FIG. 5 are the same as those described in the method of providing interpersonal services using fingerprints through FIGS. 1 to 4 above, or from the described content. Since it can be easily inferred, the following description will be omitted.

The method for providing electronic services using fingerprints according to an embodiment described with reference to FIG. 5 may be implemented in the form of a recording medium including instructions executable by a computer, such as an application or program module executed by a computer. can Computer readable media can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media, removable and non-removable media. Also, computer readable media may include all computer storage media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data.

The above-described method for providing electronic interpersonal service using a fingerprint according to an embodiment of the present invention is based on an application installed in a terminal (this may include a program included in a platform or operating system, etc. It may be executed, or it may be executed by an application (ie, a program) directly installed in the master terminal by a user through an application providing server such as an application store server, an application or a web server related to the corresponding service. In this sense, the above-described electronic service providing method using a fingerprint according to an embodiment of the present invention is implemented as an application (i.e., a program) that is basically installed in a terminal or directly installed by a user, and is implemented by a computer such as a terminal. It can be recorded on a readable recording medium.

The above description of the present invention is for illustrative purposes, and those skilled in the art can understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, the embodiments described above should be understood as illustrative in all respects and not limiting. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as distributed may be implemented in a combined form.

The scope of the present invention is indicated by the following claims rather than the detailed description above, and all changes or modifications derived from the meaning and scope of the claims and equivalent concepts should be construed as being included in the scope of the present invention. do.

Claims (10)

At least one user terminal for uploading the contract and requesting electronic seal; and
A reception unit that receives a contract uploaded from the at least one user terminal, a setting unit that sets to generate a fingerprint with a fingerprint generation parameter pre-established in each page constituting the contract, and generates a fingerprint related to the content of the page A content storage unit that stores the content fingerprint as a content fingerprint, a sequence storage unit that generates and stores a fingerprint related to the sequence, which is the page number of the page, as a sequence fingerprint, and writes the content fingerprint and sequence fingerprint as a page fingerprint A page storage unit, an electronic interpersonal service providing server including a legal tech unit that reviews risks after reviewing the contents of the contract with a pre-established artificial intelligence algorithm before including an electronic interperson after the contract is uploaded;
An electronic interpersonal service providing system using a fingerprint comprising a.
According to claim 1,
The electronic interpersonal service providing server,
Based on any one of the plurality of pages constituting the contract, a generation management unit that fills the order fingerprint with a null value if there is no previous page and fills the order fingerprint with a null value if there is no next page;
Electronic interpersonal service providing system using a fingerprint, characterized in that it further comprises.
According to claim 1,
The electronic interpersonal service providing server,
Based on any one page of the plurality of pages constituting the contract, the next page of the any one page is set to follow the any one page, and the previous page of the any one page is followed (Follow) a relationship connection unit that is set to;
Electronic interpersonal service providing system using a fingerprint, characterized in that it further comprises.
According to claim 1,
The electronic interpersonal service providing server,
Based on any one of the plurality of pages constituting the contract, the next page of the one page is set to point to the one page, and the one page is the next page of the one page a pointer unit configured to point to a page;
Electronic interpersonal service providing system using a fingerprint, characterized in that it further comprises.
According to claim 1,
The electronic interpersonal service providing server,
a repetition unit repeating the process of generating the page fingerprint until the last page of the contract;
Electronic interpersonal service providing system using a fingerprint, characterized in that it further comprises.
According to claim 5,
The electronic interpersonal service providing server,
an authentication unit for authenticating that the last page of the contract is reached and terminating the creation process;
Electronic interpersonal service providing system using a fingerprint, characterized in that it further comprises.
According to claim 1,
The electronic interpersonal service providing server,
a content hashing unit converting page contents including text and images of each page into a hash function and storing a hash value;
Electronic interpersonal service providing system using a fingerprint, characterized in that it further comprises.
According to claim 7,
The electronic interpersonal service providing server,
a sequence hashing unit converting the number contents of the order of the pages into a hash function and storing a hash value;
Electronic interpersonal service providing system using a fingerprint, characterized in that it further comprises.
delete According to claim 1,
The electronic interpersonal service providing server,
When creating a contract in the user terminal, the contents of the contract including the purpose, usage and contractor information are uploaded from the user terminal, and the type of contract is recommended, and when the type of contract is selected in the user terminal, the contract contents are an automatic creation unit that automatically creates;
Electronic interpersonal service providing system using a fingerprint, characterized in that it further comprises.

Images