Contract review server using artificial intelligence and operation method thereof

Abstract

The present invention relates to a contract review server using artificial intelligence and an operation method thereof and, more specifically, to a contract review server using artificial intelligence, which reviews and summarizes a contract by using artificial intelligence, provides main contract information, provides an answer to a question related to contract content, and helps correct incorrect contract content, and an operation method thereof.

Description

Contract review server using artificial intelligence and its operation method

[0001] The present invention relates to a contract review server utilizing artificial intelligence and an operating method thereof, and more particularly, to a contract review server utilizing artificial intelligence that uses artificial intelligence to review a contract, summarize it, provide key contract information, provide answers to questions related to the contract content, and help correct incorrect contract content, and to an operating method thereof.

[0002] In modern society, many people write and sign contracts. Contracts are being drafted not only between individuals but also between businesses on a regular basis.

[0003] Contracts are extremely diverse, encompassing general contracts, lease agreements, sales contracts, delivery contracts, authorization agreements, mergers and acquisitions, employment contracts, patent agreements, transfer of interest agreements, building sales contracts, and cooperative agreements. Furthermore, as technology advances and society evolves, the types of contracts are also diversifying, with new types emerging from traditional contracts. Furthermore, the content of contracts is becoming increasingly complex, challenging, and extensive.

[0004] However, after drafting the contract, the parties themselves had to personally review it to ensure it was drafted correctly, that there were no issues, that there were no missing terms, and that it was written to their advantage. This presented numerous challenges. Furthermore, when reviewing the contract themselves, they often overlooked aspects, making it difficult to feel confident in the contract once it was drafted.

[0005] Accordingly, there is a need for a contract review service that utilizes artificial intelligence to save time and money, reduce the number of people reviewing contracts, reduce errors and maintain consistency in reviews, and accurately correct errors in contracts.

[0006] Accordingly, the technical problem of the present invention is conceived from this point of view, and the purpose of the present invention is to provide a contract review server utilizing artificial intelligence that can review a contract much faster than a person can, and an operating method thereof.

[0007] In addition, it provides a contract review server utilizing artificial intelligence and its operation method that can save time in reviewing a large number of contracts.

[0008] In addition, a contract review server utilizing artificial intelligence and its operation method are provided to reduce the manpower required for contract review.

[0009] In addition, the present invention provides a contract review server utilizing artificial intelligence and its operation method that can reduce human errors in reviewing contracts containing complex and important contents.

[0010] In addition, the present invention provides a contract review server and its operation method utilizing artificial intelligence that eliminates subjective judgments that may arise from reviews performed by humans and maintains consistency by applying the same standards in reviewing contracts.

[0011] In addition, the present invention provides a contract review server utilizing artificial intelligence and its operation method that analyzes the contents of a contract, provides insightful information, and allows the contents of the contract to be modified to suit business strategies.

[0012] In order to achieve the above object of the present invention, a contract review server utilizing artificial intelligence is provided, which includes an input unit for a user to input a contract and information related thereto into a server; an output unit for outputting information regarding the contract contents of the contract from the server; a communication unit for communicating information inside and outside the server; a storage unit for storing information generated by the server; a control unit for controlling all operations of the server; a memory unit for executing various programs and storing accompanying data; and a learning processor for learning and generating a contract contents separation and major item matching model using an artificial neural network, wherein the memory unit includes a contract analysis unit for analyzing the structure or contents of the contract; a summary information provision unit for summarizing and providing the contents of the contract; a checklist answer unit for providing answers to checklist questions; a contract review unit for reviewing the contract; a contract storage unit for storing contracts reviewed by the review unit; and a model storage unit for storing the contract contents separation and major item matching model that is being learned or has been learned through the learning processor.

[0013] According to the present invention, by utilizing artificial intelligence to review contracts, contracts can be reviewed much faster than when performed by humans, thereby improving work efficiency.

[0014] Additionally, it can improve work efficiency by saving time when reviewing a large number of contracts.

[0015] Additionally, by utilizing artificial intelligence to review contracts, costs can be reduced by reducing the number of people required to review contracts.

[0016] Additionally, it can improve accuracy by reducing human errors in reviewing contracts containing complex and important content.

[0017] Additionally, AI can eliminate subjective judgments that may arise from human reviews and maintain consistency in reviews by applying the same standards when reviewing contracts.

[0018] Additionally, AI can be used to analyze contract content and provide insightful information that can be used to modify contract content to suit business strategies.

[0019] However, the effects of the present invention are not limited to the above effects, and can be expanded in various ways without departing from the spirit and scope of the present invention.

[0020] FIG. 1 is a block diagram of a contract review server utilizing artificial intelligence according to one embodiment of the present invention.

[0021] FIG. 2 is a diagram showing a process in which the contents of an uploaded contract according to one embodiment of the present invention are separated and matched by major items.

[0022] FIG. 3 is a diagram showing a process for providing summary information and answers to a checklist for major items according to one embodiment of the present invention.

[0023] FIG. 4 is a flowchart showing a process of separating the contents of a contract by using a semantic unit analyzer in a contract analysis unit according to one embodiment of the present invention.

[0024] FIG. 5 is an exemplary diagram showing a process of separating the contract contents of a contract by utilizing a semantic unit analyzer in a contract analysis unit according to one embodiment of the present invention.

[0025] FIG. 6 is a drawing showing a process of answering a checklist question in a checklist answer section according to one embodiment of the present invention.

[0026] FIG. 7 is a drawing showing a script written to answer checklist questions in a checklist answer section according to one embodiment of the present invention.

[0027] FIG. 8 is an exemplary diagram showing a process of providing answers to checklist questions in a checklist answer section according to one embodiment of the present invention.

[0028] FIG. 9 is a diagram showing a process of reviewing basic information of a contract written using artificial intelligence in a contract review unit according to one embodiment of the present invention.

[0029] FIG. 10 is a diagram showing a process of reviewing the provisions of a contract written using artificial intelligence in a contract review unit according to one embodiment of the present invention.

[0030] FIG. 11 is a drawing showing a process of reviewing notes of a modified contract using artificial intelligence in a contract review unit according to one embodiment of the present invention.

[0031] FIG. 12 is a drawing showing a process of storing a contract written using artificial intelligence in a contract storage unit according to one embodiment of the present invention.

[0032] FIG. 13 is a diagram showing a process of searching for contract contents stored in a contract storage unit according to one embodiment of the present invention.

[0033] Hereinafter, preferred embodiments of the present invention will be described in more detail with reference to the drawings.

[0034] The present invention is susceptible to various modifications and takes various forms. Specific embodiments are illustrated in the drawings and described in detail herein. However, this is not intended to limit the present invention to specific disclosed forms, but rather to encompass all modifications, equivalents, and alternatives falling within the spirit and technical scope of the present invention.

[0035] FIG. 1 is a block diagram of a contract review server utilizing artificial intelligence according to one embodiment of the present invention.

[0036] Referring to FIG. 1, the contract review server (100) utilizing artificial intelligence of the present invention is installed and operates on a connection terminal (10) such as a personal PC at home or a company, and is composed of an input unit (110), an output unit (120), a communication unit (130), a storage unit (140), a control unit (150), a memory unit (160), and a running processor (230).

[0037] The input unit (110) inputs a contract and related information that the user wishes to review through a connection terminal (10), the output unit (120) outputs information on the contract contents of the reviewed contract from the server, the communication unit (130) communicates information inside and outside the server, the storage unit (140) stores information generated in the server, the control unit (150) controls all operations of the server, the memory unit (160) executes various programs and stores data accompanying them, and the learning processor (220) utilizes an artificial neural network to learn a contract contents separation and main item matching model and generates a contract contents separation and main item matching model.

[0038] In addition, the memory unit (160) includes a contract analysis unit (170), a summary information provision unit (180), a checklist response unit (190), a contract review unit (200), a contract storage unit (210), and a model storage unit (220). A detailed description of each component will be provided below.

[0039] FIG. 2 is a diagram showing a process in which the contents of an uploaded contract according to one embodiment of the present invention are separated and matched by major items.

[0040] Referring to Fig. 2, the person in charge accesses the contract review server (100) utilizing artificial intelligence through the connection terminal (10) and uploads a contract to the contract review server (100) utilizing artificial intelligence. The contract analysis unit (170) analyzes the structure and content of the uploaded contract to separate the contract content into semantic units, matches the contract content of the separated semantic units with major items, and organizes the contract content of the separated semantic units by major items.

[0041] For example, if the contract is a commodity transaction contract, the main items that make up the contract may be purpose and definition, contract period, obligations of the consignor, obligations of the consignee, confidentiality, compensation for damages, force majeure, termination, performance guarantee, and insurance, and the contract contents of the contract are separated and organized by the main items mentioned above.

[0042] The main items that match the contract contents of the separated semantic units are stored in the storage unit (140) according to the type and contents of the contract. If the main items that match the contract contents of the separated semantic units are not in the storage unit (140), the contract analysis unit (170) analyzes the contract contents of the separated semantic units to generate main items corresponding to the contract contents of the analyzed semantic units, and matches the contract contents of the separated semantic units with the generated main items. In addition, the generated main items are stored in the storage unit (140) and are used when matching the main items with the contract contents of other separated semantic units.

[0043] The contract analysis unit (170) provides a 'contract display unit' through the screen of the person in charge's access terminal (10) so that the person in charge can check the uploaded contract, and provides a 'contract main item display unit' so that the person in charge can check the main items and the contract contents corresponding to the main items.

[0044] The 'Contract Display Section' and the 'Contract Main Item Display Section' are displayed simultaneously on the person in charge's access terminal screen, so that the person in charge can simultaneously check the contract provided in the 'Contract Display Section' and the main items provided in the 'Contract Main Item Display Section' and the contract contents corresponding to the main items on a single screen.

[0045] When the person in charge selects the main items provided in the 'Contract Main Item Display', the contract contents corresponding to the selected main items are displayed in color in the 'Contract Display', so the person in charge can intuitively check the contract contents corresponding to the selected main items through the 'Contract Display'.

[0046] FIG. 3 is a diagram showing a process for providing summary information and answers to a checklist for major items according to one embodiment of the present invention.

[0047] Referring to Figure 3, the summary information provision unit (180) analyzes and summarizes the separated contract contents matched with key items to provide summary information. Additionally, the checklist response unit (190) provides answers to checklist questions.

[0048] The person in charge can check the contract contents through the 'Contract Display Section' and check the summary information on major items and answers to checklist questions through the 'Summary Information and Checklist Answer Provision Section'.

[0049] Additionally, the ‘contract display section’ and the ‘summary information and checklist response provision section’ are displayed simultaneously on the screen of the person in charge’s access terminal (10).

[0050] Additionally, when the person in charge selects the summary information and checklist response displayed in the 'Summary Information and Checklist Response Provision Section', the contract contents corresponding to the selected summary information and checklist response are displayed in color in the 'Contract Display Section'.

[0051] For example, if the main item extracted by the contract analysis department (170) is 'contract period', it provides summary information on 'contract start date', 'contract end date', 'contract extension date', etc. related to the 'contract period', and provides checklist response information on 'notification when contract period is set to 3 years or more', 'presence or absence of automatic extension', etc.

[0052] FIG. 4 is a flowchart showing a process of separating the contents of a contract by using a semantic unit analyzer in a contract analysis unit according to one embodiment of the present invention.

[0053] Referring to FIG. 4, a person in charge uploads a contract in a file format (e.g., docx, doc, pdf file format) to a contract review server (100) using artificial intelligence (S1), the uploaded contract is converted into a text document (S2), a semantic unit analyzer analyzes the contract content to separate the contract content into semantic units (S3), and a contract analysis unit (170) matches the contract content of the separated semantic units with major items, and organizes the contract content of the separated semantic units by major items (S4). At this time, in step S4, the contract analysis unit (170) divides major items into common items and specific items and matches them with the contract content of the separated semantic units. Common items are items that are common regardless of the type and content of the contract, and specific items are items that vary depending on the type and content of the specific contract. For example, common items may be parties, jurisdiction, and governing law, and specific items may be the construction period and construction site of a construction contract.

[0054] FIG. 5 is an exemplary diagram showing a process of separating the contract contents of a contract by utilizing a semantic unit analyzer in a contract analysis unit according to one embodiment of the present invention.

[0055] Referring to Figure 5, the contract analysis unit (170) identifies the structure of the contract, analyzes the contract content, and separates the contract content using a semantic unit analyzer. The contract content, separated into semantic units, may be paragraphs, clauses, or articles constituting the contract, and may be sentences constituting said paragraphs, clauses, or articles.

[0056] The contract analysis unit (170) identifies and analyzes the contract's content and structure. It analyzes the beginning and end of each paragraph, clause, or clause, thereby separating the entire contract into paragraphs, clauses, or clauses. The semantic unit analyzer then analyzes the meaning of each separated paragraph, clause, or clause, separating the paragraphs, clauses, or clauses into semantic units.

[0057] The contract analysis unit (170) analyzes the beginning and end of paragraphs, clauses or provisions that make up a contract, and recognizes line breaks, spacing, indentation, numbers, and the character length of clauses in cases where there are formats such as numbers or symbols, and can separate paragraphs, clauses or provisions into meaningful units according to the language and type of the contract.

[0058] If the contract analysis unit (170) has difficulty in identifying the beginning and end of a paragraph, clause, or clause in the contract, or has difficulty in distinguishing paragraphs, clauses, or clauses, the contract analysis unit (170) separates all sentences constituting the contract into sentence units, and the semantic unit analyzer analyzes the semantic content of the separated unit sentences, and rearranges the separated unit sentences containing similar semantic content by combining them. In addition, the contract analysis unit (170) analyzes the semantic content of the rearranged unit sentences, and matches the semantic content of the rearranged unit sentences with the main items.

[0059] At this time, an encoder model such as BERT (Bidirectional Encoder Representations from Transformers) or a decoder model such as GPT (Generative Pretrained Transformer) can be used to rearrange the separated unit sentences from the entire contract into unit sentences with similar meaning.

[0060] As in Example 1 or Example 2, the contract analysis unit (170) recognizes specific characters (e.g., '제', '조', '항'), patent numbers (e.g., '2'), and the sum of specific characters and numbers (e.g., 'Article 2', 'Clause 2') that distinguish clauses of the contract, and recognizes specific symbols '()' or '[]' that follow or appear after the characters or numbers, and recognizes the text of 'confidential information' or 'product supply' within the specific symbols. If the text of 'confidential information' or 'product supply' within the recognized specific symbols is included in the main items previously stored in the storage unit (140), the semantic unit analyzer recognizes the text of 'confidential information' or 'product supply' as the main item, and separates the paragraphs or sentences that begin with a line break, space, or indentation after the text within the recognized specific symbols into the contract contents corresponding to the main items.

[0061] Meanwhile, if the text within the above-described specific symbol is not included in the main items already stored in the storage unit (140), the text within the above-described specific symbol is created as a new main item and stored in the storage unit (140), and is used when separating the contract contents of other clauses by matching them with the main items.

[0062] Additionally, if there is no text that can be recognized as a main item, the semantic unit analyzer analyzes the content of sentences that constitute a paragraph or clause, matches the main items corresponding to the contract content of the sentences that constitute the paragraph or clause in the storage unit (140), and separates the sentences that constitute the paragraph or clause using the matched main items. At this time, the semantic unit analyzer may utilize an encoder model such as BERT or a decoder model such as GPT when analyzing the content of the sentences that constitute the paragraph or clause.

[0063] Additionally, if a paragraph or clause consists of multiple sentences, the paragraph or clause is divided into sentences, and each sentence is separated by matching the main item. The separated sentences can be reordered and stored by main item, or they can be reordered and stored by semantic unit.

[0064] This allows the manager to search for contractual sentences that correspond to key items or semantic units. For example, if the manager searches for "Reason for Termination" as a key item, contractual sentences regarding termination from Article 14 (Confidentiality) and other sections, in addition to Article 12 (Termination and Cancellation of Contract), can be retrieved.

[0065] In addition, as described above, the process of analyzing the structure and content of the contract in the contract analysis unit (170), separating the contract content into semantic units using a semantic unit analyzer, and matching the separated semantic unit contract content with major items can be performed more quickly and accurately by using artificial intelligence to create a learning model and using the created learning model to perform the above process.

[0066] The contract review server (100) utilizing artificial intelligence includes a learning processor (200) and a model storage unit (220) to generate a learning model utilizing artificial intelligence. Furthermore, the storage unit (140) may serve as a big data storage database for artificial intelligence learning, and a separate big data storage database may also be provided.

[0067] The model storage unit (220) is included in the memory unit (160), and the model storage unit (220) stores the contract content separation and main item matching model that is being learned or learned through the learning processor (230).

[0068] The learning processor (220) trains a contract content separation and major item matching model using an artificial neural network, and generates a contract content separation and major item matching model. In addition, when a user uploads a contract to a contract review server (100) utilizing artificial intelligence, the contract content separation and major item matching model analyzes the structure and content of the contract based on the contract, separates the contract content into semantic units using a semantic unit analyzer, matches the separated semantic unit contract content with major items, and the separated semantic unit contract content matched with major items is rearranged and stored in a storage unit (140).

[0069] The process in which the learning processor (220) learns the contract content separation and main item matching model by utilizing an artificial neural network is the same as the process in which the contract analysis unit (170) analyzes the structure and content of the contract, separates the contract content into semantic units by utilizing a semantic unit analyzer, matches the separated semantic unit contract content with main items, and rearranges the separated semantic unit contract content matched by main items, and the learning processor (220) repeatedly learns the above process through the contract content separation and main item matching model, thereby completing the contract content separation and main item matching model.

[0070] The artificial neural network utilized by the running processor (230) may include an artificial neural network (ANN) implemented by hardware, software, or a combination thereof, and may include a convolutional neural network (CNN) trained by deep learning, but is not limited thereto and may include various forms of artificial neural networks.

[0071] The running processor (230) utilizes data in the storage unit (140) to train a contract content separation and major item matching model, and all data for which training is completed are updated and stored in the storage unit (140), so that the accuracy of the data is continuously improved.

[0072] FIG. 6 is a drawing showing a process of answering a checklist question in a checklist answer section according to an embodiment of the present invention, and FIG. 7 is a drawing showing a script written to answer a checklist question in a checklist answer section according to an embodiment of the present invention.

[0073] Referring to Figure 6, when a person in charge accesses a contract review server (100) utilizing artificial intelligence and asks a question about a checklist, the checklist answer section (190) provides an answer to the question about the checklist.

[0074] The checklist response unit (190) analyzes checklist questions, extracts key items related to the questions, analyzes and organizes contractual content into separate semantic units matching the key items, and provides the resulting information to the person in charge. If the checklist question is one that cannot be extracted as a key item, the generative AI can provide an answer. However, if the checklist question is completely unrelated to the contract, the AI ​​will respond that it cannot provide an answer.

[0075] For example, when a person in charge asks a question included in the checklist, “What are the conditions for canceling a contract?”, the checklist response section (190) determines whether the question is related to the contract (Sa1), and if it is related to the contract, analyzes the content of the question and extracts key items related to the question (Sa2), and analyzes and organizes the contract content separated into key items to provide an answer (Sa3).

[0076] Referring to FIG. 7, the checklist response unit (190) can create a response document (hereinafter referred to as a "checklist response document") for "how to answer checklist questions" in order to answer checklist questions. The checklist response document includes descriptions of major items, company policies, a list of questions, and contractual terms of the contract. When a person in charge asks a question about the checklist, the checklist response unit (190) provides an answer to the person in charge by referring to a checklist response document created in advance. The checklist response document is created for each major item, and is continuously updated and stored in the storage unit (140) as company policies, question lists, and contractual terms change.

[0077] A question can be a combination of words, such as "contract termination," or a question-like sentence, such as "What are the conditions for contract termination?" If the question is a combination of words, such as "contract termination," the system will provide an answer related to the relevant item. If the question is a question-like sentence, the system will analyze the sentence to match the most appropriate item and provide an answer related to the matched item. Furthermore, the system can analyze the question and provide an answer by referencing a pre-written checklist document.

[0078] Before the checklist response section (190) provides answers to checklist questions, the response can be reviewed for the final time. For example, if the person in charge asks a question about "compensation for damages," the checklist response section (190) matches "compensation for damages" as a key item, analyzes and organizes the contract content into separate semantic units for the key items of "compensation for damages," and provides the information to the person in charge.

[0079] However, before providing a response in the checklist response section (190), the final separate contract content to be provided is reviewed, and if it contains contract content unrelated to 'compensation for damages', the contract content unrelated to 'compensation for damages' is excluded and a response is provided to the person in charge.

[0080] Meanwhile, the person in charge may revise the contract if the answers to the checklist questions provided in the checklist response section (190) are incorrect. For example, if the answers to the checklist questions conflict with company policy, the person in charge would revise the contract to comply with company policy.

[0081] If the contract date, contract period, or other information is incorrect, the contract manager will revise the contract to correct the date, period, and other details. When revising a contract, any changes to the original text will be marked with a strikethrough (e.g., red underline) or highlighted in color, allowing the contract manager to intuitively identify the portions of the original contract that have been modified. Once the contract manager has revised the original text and confirmed the changes, the revised content will be reflected in the contract.

[0082] FIG. 8 is an exemplary diagram showing a process of providing answers to checklist questions in a checklist answer section according to one embodiment of the present invention.

[0083] Referring to Figure 8, in Figure (a), when the person in charge asks the checklist question, “Please review the compensation for damages,” the checklist response section (190) shows the related provisions, Articles 5 and 7, and summarizes the contract contents of the related provisions, Articles 5 and 7, at the bottom.

[0084] In (b), when the person in charge asks the checklist question, “Is the contract amount well aligned with the guidelines?”, the checklist response section (190) shows the relevant provisions, Articles 5 and 7, summarizes the contract contents of the relevant provisions, explains the problematic contract contents, explains the reason why the contract contents should be revised, and shows the revised contract contents of the provisions at the very bottom.

[0085] In (c), if the person in charge asks a question unrelated to the contract, a response is provided stating that the question is unrelated to the contract.

[0086] FIG. 9 is a drawing showing a process of reviewing basic information of a contract written using artificial intelligence in a contract review unit according to an embodiment of the present invention, and FIG. 10 is a drawing showing a process of reviewing provisions of a contract written using artificial intelligence in a contract review unit according to an embodiment of the present invention.

[0087] Referring to FIGS. 9 and 10, the contract review unit (200) can review contracts drafted using artificial intelligence. The person in charge can check the drafted contract in the "contract display unit" on the screen of the access terminal (10), and can proceed with the contract review process using artificial intelligence in the "artificial intelligence review unit."

[0088] The contract review department (200) can utilize artificial intelligence to automatically review contracts, utilizing a contract content separation and key item matching model that has already been trained.

[0089] The person in charge begins reviewing the contract using AI by clicking "AI Review" in the "AI Review Department." The AI-powered contract review process involves reviewing basic information and reviewing clauses.

[0090] The contract review department (200) utilizes a contract content separation and key item matching model to review contract content related to the basic information of the contract. It matches contract content related to basic information requiring review by the person in charge by key item. It then summarizes the contract content related to the basic information matched by key item through the "AI Review Department" and provides it to the person in charge. The person in charge then verifies the review items provided by the "AI Review Department" in a summary by key item.

[0091] When the person in charge selects key items for review in the "AI Review Department," the contract content corresponding to the selected items is highlighted in color in the "Contract Display Section." This allows the person in charge to identify which part of the contract they are reviewing. The person in charge reviews the contract content corresponding to the key items and, if necessary, immediately modifies the contract. The modified contract content is updated in the contract, and the person in charge can view the updated contract content in the "Contract Display Section."

[0092] The Contract Review Department (200) utilizes a contract content separation and key item matching model to review the contract's terms and conditions. The Contract Review Department (200) reviews the contract's terms and conditions and provides the person in charge with information to check regarding the contract's contents.

[0093] For example, the contract review department (200) reviews the terms of the contract and informs the person in charge of the review regarding the contract period. When reviewing the terms, the contract review department (200) can refer to key items (e.g., "contract period") contained within specific characters (e.g., "( )") following the number of the clause (e.g., "Article 2") and analyze the sentences contained within the clause.

[0094] The contract review department (200) recognizes the text corresponding to the main item called 'contract period' contained within specific characters () to confirm that the clause is related to the contract period, and if it does not recognize the text corresponding to the main item, it analyzes the sentences contained in the clause to confirm that the clause is related to the contract period.

[0095] If the contract review department (200) confirms that the relevant clause relates to a key item within the contract period, it locates the year, month, and day corresponding to the contract period and verifies the contract period. After confirming the contract period, the contract review department (200) informs the person in charge of the items to be checked within the contract period through the "AI Review Department." At this time, the contract review department can summarize not only the contract period but also the contractual terms related to the contract period and provide the person in charge with the information.

[0096] At this time, the text corresponding to the contract period is colored and displayed in the 'Contract Display Section' so that the person in charge can check it, and the 'AI Review Section' displays the items that the person in charge must check so that the person in charge can check them.

[0097] For example, the contract review department (200) displays the review contents regarding ‘presence or absence of automatic extension’ as an item to be checked by the person in charge in relation to the contract period in the ‘artificial intelligence review department.’

[0098] The contract review department (200) can select key items to display for review through the "AI review department." The contract review department (200) can select key items to display in the "AI review department" based on the type or category of the contract. Furthermore, the contract review department (200) can review and notify the company of whether the review content for key items displayed in the "AI review department" conforms to the company's policies.

[0099] FIG. 11 is a drawing showing a process of reviewing notes of a modified contract using artificial intelligence in a contract review unit according to one embodiment of the present invention.

[0100] Referring to Figure 12, the contract review department (200) locates the contract section where notes were written and notifies the person in charge of the notes through the "Memo Review Department." The person in charge can then use the "Memo Review Department" to confirm who wrote the notes, when, and what kind of notes were written. The notes are displayed in chronological order, and the person in charge can filter the notes by author and time.

[0101] The person in charge can also review revised contract content through the "Contract Display" section. Deleted contract text is marked with a red line, and revised or added text is highlighted in color.

[0102] When the person in charge selects the contract text that has been deleted, modified, or added in the 'Contract Display Section', he or she can check the memo corresponding to the contract text selected in the 'Memo Review Section'. When the person in charge selects a memo in the 'Memo Review Section', he or she can check the contract text corresponding to the memo selected in the 'Contract Display Section'.

[0103] FIG. 12 is a drawing showing a process of storing a contract written using artificial intelligence in a contract storage unit according to one embodiment of the present invention.

[0104] Referring to Figure 12, the contract storage unit (210) can store contracts written using artificial intelligence. The person in charge can check the previously written contract in the "contract display unit" on the screen of the access terminal (10), and can proceed with the process of saving the contract using artificial intelligence in the "artificial intelligence storage unit."

[0105] The contract storage unit (210) can utilize a contract content separation and main item matching model that has already been learned using artificial intelligence to automatically store the contract.

[0106] The person in charge uploads the completed electronic document contract to a contract review server (100) utilizing artificial intelligence, and the uploaded electronic contract is converted into text data using OCR (Optical Character Recognition) technology so that it can be searched and stored as text.

[0107] The contract storage unit (210) uses artificial intelligence to review the structure and contract contents of a contract converted into text data, separates the contract contents into meaning units, matches the contract contents separated into meaning units with main items, and rearranges and stores the contract contents of meaning units matched with main items.

[0108] The person in charge reviews the contract contents, organized by key items and stored in the "AI storage." Key items in the stored contract contents may include the contracting party, the contracting party, the contract start date, the contract end date, key contract items, and key provisions. The person in charge conducts a final review of the contract contents organized by key items and, if there are no issues, saves the finalized contract contents.

[0109] FIG. 13 is a diagram showing a process of searching for contract contents stored in a contract storage unit according to one embodiment of the present invention.

[0110] Referring to Figure 13, the person in charge can also use the key items identified in the contract storage unit (210) as search terms when searching for contract content. The person in charge can search for contract content by entering the key items as search terms in the "Contract Search Window." In this case, not only the contract content but also information such as the contract name, contract category, contract counterparty, and contract start date are searched, allowing the person in charge to check the above information. Furthermore, the person in charge can download the searched contract content list in an Excel file format and save it on their access terminal (10).

[0111] Although the present invention has been described with reference to the above embodiments, it will be understood by those skilled in the art that various modifications and changes can be made to the present invention without departing from the spirit and scope of the present invention as set forth in the claims below.

[0112] [Explanation of symbols]

[0113] 10: Connection terminal

[0114] 100: Contract Review Server Using AI

[0115] 110: Input section

[0116] 120: Output section

[0117] 130: Communications Department

[0118] 140: Storage

[0119] 150: Control unit

[0120] 160: Memory section

[0121] 170: Contract Analysis Department

[0122] 180: Summary Information Provider

[0123] 190: Checklist Answer Sheet

[0124] 200: Contract Review Department

[0125] 210: Contract storage

[0126] 220: Model storage

[0127] 230: Running Processor

Claims

1. Input section where users input contracts and related information into the server; An output section that outputs information regarding the contract contents of the above contract from the server; A communication unit that communicates information inside and outside the server; A storage unit that stores information generated by the server; A control unit that controls all operations of the server; In a contract review server utilizing artificial intelligence, including a memory unit that executes various programs and stores accompanying data; and a learning processor that learns and generates a contract content separation and major item matching model using an artificial neural network, The above memory unit includes a contract analysis unit that analyzes the structure or content of the contract; Summary information provision section that summarizes the contents of the above contract; A checklist answer section providing answers to the checklist questions; Contract Review Department that reviews the above contract; A contract review server utilizing artificial intelligence, comprising a contract storage unit for storing contracts that have been reviewed by the above review unit; and a model storage unit for storing the contract content separation and main item matching model that is being learned or learned through the above learning processor.

2. In paragraph 1, The above contract analysis unit is a contract review server that utilizes artificial intelligence to analyze the structure or content of the above contract, separate the contract content of the above contract into meaningful units, and match the contract content of the separated meaningful units with major items.

3. In paragraph 2, The above contract analysis unit is a contract review server that utilizes artificial intelligence to provide the contract through a contract display section on the user's access terminal screen, and to provide the main items and the contract contents corresponding to the main items through a contract main item display section, so that the user can simultaneously check the contract, the main items, and the contract contents corresponding to the main items.

4. In paragraph 3, A contract review server utilizing artificial intelligence that displays the contract contents corresponding to the selected main items in color in the contract display section when the user selects the main items provided in the main items display section of the contract.

5. In paragraph 2, The above main items that match the contract contents of the above separated meaning units are stored in the storage unit according to the type and contents of the above contract, and are a contract review server utilizing artificial intelligence.

6. In paragraph 2, The above main items matching the contract contents of the above separated meaning units are divided into common items that are common regardless of the type and content of the above contract and specific items that vary depending on the type and content of the above contract, and a contract review server utilizing artificial intelligence.

7. In paragraph 5, If the contract content of the separated meaning unit is not stored in the storage unit, the contract analysis unit analyzes the contract content of the separated meaning unit and creates a contract review server using artificial intelligence that matches the contract content of the separated meaning unit.

8. In paragraph 2, The contract content of the above separated meaning unit is a paragraph, article or clause constituting the above contract, or a contract review server using artificial intelligence that is a sentence constituting the above paragraph, article or clause.

9. In paragraph 8, If the contract analysis unit cannot distinguish the paragraphs, clauses or articles that constitute the contract, the contract analysis unit divides the entire sentences that constitute the contract into sentence units, analyzes the semantic content of the divided sentence units, and rearranges the divided sentence units that contain similar semantic content by combining them. A contract review server utilizing artificial intelligence.

10. In paragraph 9, The above analysis unit analyzes the semantic content of the separated unit sentences, and uses an encoder model such as BERT (Bidirectional Encoder Representations from Transformers) or a decoder model such as GPT (Generative Pretrained Transformer) to rearrange the separated unit sentences containing similar semantic content. This is an artificial intelligence-based contract review server.

11. In paragraph 9, The above contract analysis unit analyzes the semantic content of the rearranged unit sentences, and is a contract review server that utilizes artificial intelligence to match the analyzed semantic content with the above main items.

12. In paragraph 2, The above checklist response section is a contract review server utilizing artificial intelligence that analyzes the above checklist questions, extracts key items related to the above checklist questions, analyzes and organizes the contract contents of the separated semantic units matching the key items, and provides them to the user.

13. In paragraph 1, The above learning processor is a contract review server that utilizes artificial intelligence to repeatedly learn the process of analyzing the structure and content of the contract, separating the contract content of the contract into semantic units, matching the contract content of the separated semantic units with major items, and reorganizing the contract content of the separated semantic units matched by major items through the contract content separation and major item matching model.

14. In paragraph 13, The above contract review department uses the contract content separation and main item matching model to review the contract content related to the basic information of the above contract, matches the contract content related to the basic information that requires the user's review by main item, and provides the contract content related to the basic information matched by main item to the user by summarizing it. A contract review server utilizing artificial intelligence.

Images