A contract content verification method and system, an electronic device, and a storage medium

Abstract

This application provides a contract content verification method, system, electronic device, and storage medium, relating to the field of data processing, which can improve the accuracy of verification results obtained by current contract content verification technologies. The method is applied to an electronic device, which stores a first knowledge graph corresponding to at least one contract type in at least one domain. The first knowledge graph is used to characterize the relationship between the contract content, source basis, and risk level corresponding to a contract type. The method includes: the electronic device acquiring a target contract; constructing a second knowledge graph corresponding to the target contract based on the target first knowledge graph matching the contract type; the second knowledge graph being used to characterize the relationship between the target contract content, source basis, and risk level; and the electronic device acquiring the verification result of the target contract based on the second knowledge graph, the verification result including at least the content risk item of the target contract.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of data processing, and in particular to a contract content verification method and system, an electronic device, and a storage medium. BACKGROUND

[0002] Under the background of the rapid development of digital economy, as the core document of regulating the rights and obligations of transaction subjects and bearing commercial agreement, the verification quality of the contract is directly related to the risk prevention and control and legal rights protection of enterprises. Therefore, efficient and accurate verification technology is crucial for the verification of contract content.

[0003] At present, the contract content verification technology has been freed from pure manual dependence, and gradually introduces knowledge graph for quality verification of contract content.

[0004] However, the accuracy of the verification result obtained by the current contract content verification technology is low. SUMMARY

[0005] The present application provides a contract content verification method and system, an electronic device, and a storage medium, which improves the accuracy of the verification result obtained by the current contract content verification technology.

[0006] In a first aspect, the present application provides a contract content verification method, applied to an electronic device, and the electronic device stores at least one first knowledge graph corresponding to at least one contract type in at least one field, and the first knowledge graph is used to represent the relationship between the contract content corresponding to a contract type, the source basis of the contract content, and the risk level of the contract content.

[0007] The method comprises: The electronic device obtains a target contract, constructs a second knowledge graph corresponding to the target contract based on a target first knowledge graph matched with the contract type of the target contract, and the second knowledge graph is used to represent the relationship between the target contract content of the target contract, the source basis of the target contract content, and the risk level of the target contract content. The electronic device obtains the verification result of the target contract based on the second knowledge graph, and the verification result at least includes the content risk item of the target contract.

[0008] Based on the first aspect, a target first knowledge graph corresponding to the contract type is selected based on the target contract and contract type. Compared with a general knowledge graph, using the target first knowledge graph avoids redundant matching of irrelevant content and improves the adaptability and accuracy between the target contract and the knowledge graph. Based on the target first knowledge graph that is more closely matched to the target contract, the constructed second knowledge graph corresponding to the target contract has higher accuracy and matching. The verification results obtained based on the second knowledge graph corresponding to the target contract have a higher relevance to the target contract than the verification results obtained based on a general knowledge graph or a target first knowledge graph of the corresponding type, thus resulting in higher accuracy of the verification results.

[0009] In one possible implementation, the above method also includes: The electronic device decomposes the constraints corresponding to the contract content in the contract verification basis, obtaining knowledge atoms corresponding to each constraint. Each knowledge atom includes an identifier, the corresponding constraint, the applicable scenario for the constraint, and the verification rule for the constraint. Based on the knowledge atoms corresponding to each constraint, the electronic device constructs a first knowledge graph.

[0010] In another possible implementation, the aforementioned electronic device constructs a first knowledge graph based on the knowledge atoms corresponding to each constraint, including: The electronic device extracts entities from the contract verification basis based on each knowledge atom, obtaining multiple entity nodes. Each entity node includes at least the constraint condition corresponding to each knowledge atom, the constraint object of the constraint condition corresponding to the knowledge atom, the source basis of the constraint condition corresponding to the knowledge atom, and the risk level of the constraint condition corresponding to the knowledge atom. Based on the association rules between entity nodes, the electronic device obtains the relationship edges between pairs of entity nodes, constructing a first knowledge graph.

[0011] In another possible implementation, the above method also includes: The electronic device acquires contract verification criteria based on a preset acquisition cycle. If new contract verification criteria are acquired, the electronic device updates the first knowledge graph based on the new contract verification criteria to obtain the updated first knowledge graph.

[0012] In another possible implementation, the aforementioned electronic device decomposes the constraints corresponding to the contract content in the contract verification basis to obtain knowledge atoms corresponding to each constraint, including: The electronic device decomposes the constraints corresponding to the contract content in the contract verification basis, obtains the initial knowledge atoms corresponding to each constraint, performs data verification on multiple initial knowledge atoms, and takes the initial knowledge atoms that pass the data verification as knowledge atoms.

[0013] In another possible implementation, the aforementioned electronic device performs data verification on multiple initial knowledge atoms, and uses the initial knowledge atoms that pass the data verification as knowledge atoms, including: The electronic device performs integrity verification on multiple initial knowledge atoms and normative verification on the initial knowledge atoms based on an expert rule base. Thus, the electronic device can use the initial knowledge atoms that pass both integrity and normative verification as knowledge atoms.

[0014] In another possible implementation, the above method also includes: Electronic devices store knowledge atoms in a knowledge atom database and create type indexes corresponding to different types of knowledge atoms based on their types.

[0015] The aforementioned electronic device constructs a first knowledge graph based on knowledge atoms corresponding to multiple constraints, including: The electronic device retrieves knowledge atoms corresponding to multiple constraints from the knowledge atom database based on the type index of the knowledge atoms corresponding to multiple constraints, and constructs the first knowledge graph.

[0016] In another possible implementation, the aforementioned electronic device constructs a second knowledge graph corresponding to the target contract based on a first knowledge graph that matches the contract type of the target contract, including: The electronic device extracts the target entity from the target contract. The target entity includes at least subject-type variables, object-type variables, performance-type variables, or liability-type variables. The electronic device obtains a first knowledge graph that matches the contract type of the target contract. From the first knowledge graph, the electronic device extracts entity nodes and relationship edges that match the target entity, and obtains a second knowledge graph corresponding to the target contract.

[0017] In another possible implementation, the aforementioned electronic device extracts the target entity from the target contract, including: The electronic device inputs the target contract into a preset field extraction model for field extraction to obtain the target entity. Alternatively, the electronic device extracts the target entity from the target contract using a regular expression matching algorithm.

[0018] In another possible implementation, after the aforementioned electronic device extracts the target entity from the target contract, the method further includes: The electronic device performs logical and integrity checks on the target entity according to the corresponding constraints, and marks the target entity that fails the logical check or integrity check as the target entity to be verified; the verification result includes the target entity to be verified.

[0019] In another possible implementation, the first knowledge graph has a three-level structure, which includes knowledge atoms, the source basis of the constraints corresponding to the knowledge atoms, and the risk level of the constraints corresponding to the knowledge atoms. The second knowledge graph has a four-level structure, which includes the target entity, the target knowledge atoms corresponding to the target entity, the source basis of the target constraints corresponding to the target knowledge atoms, and the risk level of the target constraints corresponding to the target knowledge atoms.

[0020] In another possible implementation, the aforementioned electronic device obtains the verification result of the target contract based on a second knowledge graph, including: The electronic device retrieves the target verification template corresponding to the contract type and domain attribute from the preset verification template library based on the contract type and domain attribute of the target contract. The electronic device then fills the target verification template based on the second knowledge graph to obtain the verification result of the target contract.

[0021] In another possible implementation, the above method also includes: The electronic device sends the verification result of the target contract to the target user corresponding to the target contract through a preset method. The electronic device obtains the feedback data of the target user, updates the target verification template corresponding to the target user's verification result, stores the updated target verification template in the preset verification template library, and updates the preset verification template library.

[0022] The aforementioned electronic device, based on the contract type and domain attributes corresponding to the target contract, retrieves the target verification template corresponding to the contract type and domain attributes from a preset verification template library, including: The electronic device retrieves the target verification template corresponding to the contract type and domain attribute from the updated preset verification template library based on the contract type and domain attribute of the target contract.

[0023] In another possible implementation, the above method also includes: The electronic device sends the verification result of the target contract to the target user corresponding to the target contract through a preset method. The electronic device obtains the feedback data of the target user, and then updates the knowledge atom based on the feedback data to obtain the updated knowledge atom.

[0024] The aforementioned electronic device constructs a first knowledge graph based on the knowledge atoms corresponding to each constraint, including: The electronic device constructs the first knowledge graph based on the updated knowledge atoms corresponding to each constraint.

[0025] In another possible implementation, the above method also includes: The electronic device sends the verification result of the target contract to the target user corresponding to the target contract through a preset method. The electronic device obtains the feedback data of the target user, and then updates the parameters of the regular expression matching algorithm based on the feedback data to obtain the updated regular expression matching algorithm.

[0026] Electronic devices extract target entities from target contracts using regular expression matching algorithms, including: Electronic devices extract target entities from target contracts using an updated regular expression matching algorithm.

[0027] Secondly, embodiments of this application provide a contract review system, which includes a knowledge graph generation module and a verification result generation module.

[0028] The knowledge graph generation module is used to construct a second knowledge graph corresponding to the target contract based on the first target knowledge graph that matches the contract type of the target contract.

[0029] The verification result generation module is used to obtain the verification result of the target contract based on the second knowledge graph. The verification result includes at least the content risk items of the target contract.

[0030] Thirdly, this application provides an electronic device, which includes: a display, a processor, a communication interface, and a memory. The memory stores at least one computer program, which is loaded and executed by the processor to implement a contract content verification method or a contract content verification system.

[0031] Fourthly, this application provides a computer-readable storage medium storing at least one computer program, which is loaded and executed by a processor to implement a contract content verification method or a contract content verification system.

[0032] Fifthly, this application provides a computer program product, which includes a computer program or instructions that, when executed by a processor, implement a contract content verification method or a contract content verification system.

[0033] It should be noted that any of the possible implementations of any of the above aspects can be combined, provided that the solutions do not contradict each other. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application; Figure 2 A flowchart illustrating a contract content verification method provided in an embodiment of this application; Figure 3 This application provides a schematic flowchart of a method for constructing a first knowledge graph. Figure 4 This application provides a schematic diagram of the structure of a contract content verification system. Figure 5 This is a schematic diagram of a contract content verification device provided in an embodiment of this application. Detailed Implementation

[0036] The following describes in detail, with reference to the accompanying drawings, a method for verifying contract content provided in an embodiment of this application.

[0037] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0038] The terms "first" and "second," etc., used in the specification and drawings of this application are used to distinguish different objects or to distinguish different treatments of the same object, rather than to describe a specific order of objects.

[0039] Furthermore, the terms "comprising" and "having," and any variations thereof, used in the description of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include other steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus.

[0040] It should be noted that in the embodiments of this application, the words "exemplary" or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0041] The following explanations of some terms used in the embodiments of this disclosure are provided to facilitate understanding by those skilled in the art: (1) Knowledge graph: It is a structured semantic knowledge base with graph structure as the core. It symbolically describes the concepts, entities and their semantic relationships in the physical world through "entity-relationship-entity" triples and "entity-attribute-value" pairs, supporting machine understanding and reasoning.

[0042] (2) Contract review: This is a professional entity's comprehensive analysis and verification of the contract's terms, rights and obligations, performance risks, etc., based on the agreed terms, industry rules, transaction purpose and risk control requirements. The core is to systematically review and identify loopholes, clause defects and potential risks in the contract to ensure that the contract is legal and valid, with clear rights and responsibilities and strong enforceability, so as to ultimately achieve the transaction purpose and protect the legitimate rights and interests of the parties.

[0043] In the context of the rapid development of the digital economy, contracts, as core documents that regulate the rights and obligations of transacting parties and embody commercial agreements, are directly related to the quality of their verification, which in turn affects corporate risk control and the protection of legitimate rights and interests. Therefore, efficient and accurate verification technology is crucial for verifying contract content.

[0044] Currently, contract content verification technology has moved beyond purely manual reliance and is gradually incorporating knowledge graphs for quality verification of contract content.

[0045] However, the accuracy of the verification results obtained by current contract content verification technology is relatively low.

[0046] To address the aforementioned technical issues, the contract content verification method provided in this application involves an electronic device constructing a second knowledge graph corresponding to the target contract based on a first knowledge graph that matches the contract type of the target contract. This allows the verification result of the target contract to be obtained based on the second knowledge graph.

[0047] In this embodiment, a target first knowledge graph corresponding to the contract type is selected based on the target contract and contract type. Compared with a general knowledge graph, using the target first knowledge graph avoids redundant matching of irrelevant content and improves the adaptability and accuracy between the target contract and the knowledge graph. Based on the target first knowledge graph that is more closely matched to the target contract, the constructed second knowledge graph corresponding to the target contract has higher accuracy and matching. The verification results obtained based on the second knowledge graph corresponding to the target contract have a higher relevance to the target contract than the verification results obtained based on a general knowledge graph or a target first knowledge graph of the corresponding type, thus resulting in higher accuracy of the verification results.

[0048] The contract content verification method provided in this application can be applied to electronic devices.

[0049] Among them, electronic devices can be computing devices, terminals, servers, etc. For example, terminals can be mobile phones, tablets, computers with wireless transceiver capabilities, virtual reality (VR) terminal devices, augmented reality (AR) terminal devices, mixed reality (MR) terminal devices, terminals in industrial control, vehicle-mounted terminal devices, terminals in self-driving, terminals in assisted driving, terminals in remote medical care, terminals in smart grids, terminals in transportation safety, terminals in smart cities, terminals in smart homes, etc.

[0050] A terminal may also be referred to as terminal equipment, user equipment (UE), access terminal, vehicle-mounted terminal, industrial control terminal, remote terminal, mobile device, UE terminal equipment, wireless communication equipment, machine terminal, or UE device, etc. A terminal can be fixed or mobile.

[0051] A server can be a single physical server; or it can consist of two or more physical servers that share different responsibilities, working together to achieve the corresponding functions. In terms of server type, for example, a server can be a blade server, a high-density server, a rack server, or a tower server, etc.

[0052] Taking electronic devices as servers as an example, Figure 1This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application.

[0053] See Figure 1 , Figure 1 The electronic device 100 shown may include: a display 101, a processor 102, a memory 103, a communication interface 104, and a bus 105. The processor 102, the memory 103, and the communication interface 104 can be connected to each other via the bus 105.

[0054] The processor 102 is the control center of the electronic device. It can be a general-purpose central processing unit (CPU) or other general-purpose processors. The general-purpose processor can be a microprocessor or any conventional processor.

[0055] In this embodiment, the processor 102 can serve as the execution body of the translation method, used for translation processing between different languages.

[0056] As an example, processor 102 may include one or more CPUs, for example Figure 1 CPU 0 and CPU 1 are shown in the diagram.

[0057] The memory 103 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), disk storage medium or other magnetic storage device, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto.

[0058] In one possible implementation, the memory 103 may exist independently of the processor 102. The memory 103 can be connected to the processor 102 via a bus 105 and is used to store data, instructions, or program code. When the processor 102 calls and executes the instructions or program code stored in the memory 103, it can implement the service data acquisition method provided in the embodiments of this application.

[0059] In another possible implementation, the memory 103 can also be integrated with the processor 102.

[0060] The communication interface 104 is used for the electronic device to connect with other devices via a communication network, which may be Ethernet, radio access network (RAN), wireless local area network (WLAN), etc. The communication interface 104 may include a receiving unit for receiving data and a transmitting unit for transmitting data.

[0061] Bus 105 can be an industry standard architecture (ISA) bus, a peripheral component interconnect (PCI) bus, or an extended industry standard architecture (EISA) bus. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 1 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0062] It should be pointed out that, Figure 1 The structure shown does not constitute a limitation on the electronic device, except... Figure 1 In addition to the components shown, the electronic device may include more or fewer components than illustrated, or combine certain components, or have different component arrangements.

[0063] Figure 2 This is a flowchart illustrating a contract content verification method provided in an embodiment of this application, as shown below. Figure 2 As shown, the method includes: The electronic device stores a first knowledge graph corresponding to at least one type of contract in at least one domain.

[0064] In this embodiment, "domain" refers to a specific category with clearly defined boundaries, professional attributes, or application scenarios. For example, a domain may include the medical field, the financial field, the legal field, the consulting field, the aerospace field, the telecommunications field, etc.

[0065] A contract type refers to a specific category of contract with clearly defined terms, applicable rules, boundaries of rights and obligations, and methods of performance. For example, contract types may include sales contracts, loan contracts, guarantee contracts, lease contracts, transportation contracts, brokerage contracts, and agency contracts.

[0066] The first knowledge graph is used to represent the relationship between the contract content, the source of the contract content, and the risk level of the contract content corresponding to a certain type of contract.

[0067] It should be noted that different fields may include the same contract types, such as sales contracts, agency contracts, and loan contracts. These contract types exist in most fields and can be called general contract types.

[0068] In one feasible implementation, the first knowledge graph corresponding to a general contract type can be the same; that is, the first knowledge graph of the same type of contract in different domains is the same.

[0069] In another feasible implementation, when verifying the contract content of the same type of contract, the contract content verified in different domains is different. Therefore, the first knowledge graph corresponding to the general contract type can be different. That is, the first knowledge graph of the same type of contract in different domains is different.

[0070] In contrast to general contract types are domain-specific contract types, meaning these types of contracts have a one-to-one correspondence with specific domains. For example, transportation contracts belong specifically to the transportation and logistics domain, trust contracts and factoring contracts belong specifically to the financial domain, and medical service contracts and physical examination service contracts belong specifically to the medical domain. These contract types have specific application areas, therefore, the corresponding first knowledge graphs for these contract types are different.

[0071] S201, Target Contract for Acquiring Electronic Equipment.

[0072] The target contract is used to represent a contract whose content needs to be verified. Target contracts can exist in various forms, including electronic contracts and paper contracts.

[0073] Contract content verification is used to check and review whether the contract content complies with existing agreements, industry regulatory requirements, and transaction practice standards. The contract content to be verified includes the contracting parties, terms, rights and obligations, conditions for effectiveness, performance agreements, and liabilities for breach of contract.

[0074] In this embodiment of the application, contract content verification can also be referred to as comprehensive contract content verification or contract content compliance review.

[0075] Electronic devices can obtain target contracts through various means. For example, if the target contract is electronic, the electronic device can receive the target contract sent by the user's terminal device, or it can obtain the target contract that the user has inserted into the storage medium of the electronic device, and so on. If the target contract is a paper contract, the electronic device can scan the target contract using a camera to obtain a scanned copy of the target contract.

[0076] In some embodiments, the electronic device acquiring the target contract can be regarded as the initial target contract. The initial target contract includes contract content that needs to be verified, such as the deposit amount as a percentage of the total contract amount, whether the entity has the corresponding industry license for the transaction, whether the subject matter is legal, whether the performance method violates the mandatory provisions of the existing agreement clauses, and whether the electronic contract has a reliable electronic signature and seal.

[0077] The initial target contract also includes contract content that does not need to be verified, such as purely decorative expressions in the initial target contract like "we are willing to work together and cooperate for a long time", redundant expressions that appear repeatedly in the initial target contract, and expressions that are not related to the initial target contract itself, such as the partner's company profile, promotional descriptions of the target, and irrelevant remarks at the end of the contract.

[0078] Therefore, electronic devices do not need to verify all the contents of the initial target contract. Instead, they can first filter the contents of the initial target contract to obtain the contents that need to be verified, thereby improving the efficiency of electronic devices in verifying contract contents.

[0079] S202. The electronic device constructs a second knowledge graph corresponding to the target contract based on the first knowledge graph that matches the contract type of the target contract.

[0080] The second knowledge graph is used to characterize the relationship between the content of the target contract and the source basis of the target contract content, as well as the risk level of the target contract content.

[0081] In some embodiments, the target first knowledge graph includes the relationship between the contract content corresponding to the contract type of the target contract, the source basis of the contract content, and the risk level of the contract content. The electronic device extracts the target contract content of the target contract, and based on the target first knowledge graph, finds contract content that matches the target contract content. This allows the construction of the relationship between the target contract content and its source basis and risk level, resulting in a second knowledge graph corresponding to the target contract.

[0082] S203. The electronic device obtains the verification result of the target contract based on the second knowledge graph.

[0083] The verification results should include at least the content risk items of the target contract.

[0084] The verification results may be presented in at least one of the following formats: contract review opinion, contract review annotation draft, contract verification checklist, and second knowledge graph.

[0085] For example, a contract review opinion may include basic information about the target contract, the results of the verification of the target contract content, suggestions for modifying the target contract, and conclusive opinions on the verification.

[0086] Content risk items are used to characterize the anomalies in the contract's content. For example, content risk items can be categorized into high-risk, medium-risk, and low-risk items.

[0087] The following section introduces several methods for electronic devices to obtain the verification results of target contracts based on a second knowledge graph.

[0088] In one feasible implementation, a universal verification template is deployed in the electronic device. Based on the relationship between the contract content, source basis, and risk level of the target contract in the second knowledge graph, the electronic device fills the universal verification template with the contract content, source basis, and risk level of the target contract to obtain the verification result of the target contract.

[0089] In another feasible implementation, the electronic device deploys a pre-set verification template library, which includes verification templates corresponding to contract types and domain attributes. Based on the contract type and domain attribute of the target contract, the electronic device retrieves the target verification template corresponding to that contract type and domain attribute from the pre-set verification template library. Then, based on the relationship between the contract content, source basis, and risk level of the target contract in the second knowledge graph, the electronic device fills the target verification template with the contract content, source basis, and risk level of the target contract, thus obtaining the verification result of the target contract.

[0090] The verification templates in the preset verification template library can be generated based on historical verification results through a verification template generation model. In this embodiment, a target first knowledge graph corresponding to the contract type is selected based on the target contract and contract type. Compared with a general knowledge graph, using a target first knowledge graph avoids redundant matching of irrelevant content, improving the adaptability and accuracy between the target contract and the knowledge graph. Based on the target first knowledge graph that is more closely matched to the target contract, the constructed second knowledge graph corresponding to the target contract has higher accuracy and matching. The verification results obtained based on the second knowledge graph corresponding to the target contract have a higher relevance to the target contract than the verification results obtained based on a general knowledge graph or a target first knowledge graph of the corresponding type, thus resulting in higher accuracy of the verification results.

[0091] Figure 3 This is a schematic flowchart illustrating a method for constructing a first knowledge graph, provided as an embodiment of this application. The method includes: In this embodiment, a first knowledge graph is used to represent the knowledge graph corresponding to a certain contract type in a certain domain. The contract verification basis in this embodiment belongs to the verification knowledge of a certain contract type in a certain domain.

[0092] S301. The electronic device decomposes the various constraints corresponding to the contract content in the contract verification basis and obtains the knowledge atoms corresponding to each constraint.

[0093] Contract verification criteria refer to the content used to verify the contract content. For example, contract verification criteria include agreed clauses, industry compliance standards, and historical review data.

[0094] Constraints are used to characterize the smallest indivisible constraint in the field of contract verification that has independent verification guidance value. In other words, after further breaking down a constraint, the resulting content cannot be used as an independent constraint; it must be combined with other content to become an independent constraint.

[0095] In this embodiment, a knowledge atom is defined by a unique identifier, corresponding constraints, applicable scenarios for those constraints, and verification rules for those constraints. This defines the smallest functional unit in the contract verification field that cannot be further divided and possesses independent verification guidance value. In other words, further subdividing a single knowledge atom results in a single small functional unit that cannot serve as the basis for independent verification guidance; it must be combined with other small functional units to become the basis for independent verification guidance.

[0096] Among them, the constraints corresponding to knowledge atoms can also be called the core definitions, certified concepts, etc. of knowledge atoms.

[0097] For example, a knowledge atom could be identified as KA-001, with the constraint that the deposit ratio must not exceed 20% of the total contract amount. This constraint applies to all contracts involving deposit agreements. The verification rule for this constraint is to calculate the deposit ratio between the deposit amount and the total contract amount; a deposit ratio of 20% or less is compliant. A knowledge atom could also be identified as KA-023, with the constraint that the breach of contract liability agreement must be related to the actual loss. This constraint applies to all contracts involving breach of contract liability clauses. The verification rule for this constraint is to identify the breach of contract liability triggering conditions, compensation amount, and calculation method, and determine whether these conditions are directly related to the contract subject matter and the performance of obligations, etc.

[0098] S302. The electronic device constructs the first knowledge graph based on the knowledge atoms corresponding to each constraint condition.

[0099] In the embodiments of this application, knowledge atoms include various types of knowledge atoms, such as knowledge atoms of source basis, clause compliance, risk identification, and logic verification.

[0100] It is understandable that there is a unique correspondence between different types of knowledge atoms. That is, there is a unique correspondence between a knowledge atom of clause compliance, a unique knowledge atom of source basis, a unique knowledge atom of risk identification, and a unique knowledge atom of logic verification.

[0101] In one feasible implementation, electronic devices construct a first knowledge graph of knowledge atoms by associating relationships between different types of knowledge atoms.

[0102] In another feasible implementation, the electronic device can extract entities from the contract verification basis to obtain multiple entity nodes. Based on the verification rules of the knowledge atoms corresponding to each constraint, the electronic device obtains the association rules between entity nodes. Thus, based on the association rules between entity nodes, the electronic device can obtain the relationship edges between any two entity nodes, constructing a first knowledge graph. Here, each entity node includes at least the constraint corresponding to each knowledge atom, the constraint object of the constraint corresponding to the knowledge atom, the source basis of the constraint corresponding to the knowledge atom, and the risk level of the constraint corresponding to the knowledge atom.

[0103] An entity is used to represent a real object that can be uniquely identified and has a specific target within a certain domain. The real object includes the constraint object, the source basis, the risk level, etc.

[0104] In this embodiment of the application, the first knowledge graph includes multiple entity nodes and relationship edges between pairs of entity nodes.

[0105] For example, the entities extracted by the electronic device include the deposit amount, the 20% ratio limit, and Article XXX of the agreement. The verification rule corresponding to knowledge atom 1 is to calculate the deposit ratio between the deposit amount and the total contract amount. A deposit ratio of less than or equal to 20% is compliant. The verification rule corresponding to knowledge atom 2 is that a deposit ratio of less than or equal to 20% falls under Article XXX of the agreement. Based on the verification rule corresponding to knowledge atom 1, the electronic device can obtain the association between the entity node corresponding to the deposit amount and the entity node corresponding to the 20% ratio limit, which is "Deposit Amount - Must Comply with - 20% Ratio Limit". Based on the verification rule corresponding to knowledge atom 2, the electronic device can obtain the association between the entity node corresponding to the 20% ratio limit and the entity node corresponding to Article XXX of the agreement, which is "20% Ratio Limit - Based on - Article XXX of the Agreement".

[0106] In some embodiments, the structure of the first knowledge graph can be a three-level structure, which includes knowledge atoms, the source basis of the constraints corresponding to the knowledge atoms, and the risk level of the constraints corresponding to the knowledge atoms. Each knowledge atom includes an identifier, the corresponding constraint, the applicable scenario of the constraint, and the verification rules for the constraint.

[0107] Optionally, the structure of the first knowledge graph can also be a four-level structure, which includes knowledge atoms, the clause type of the constraint conditions corresponding to the knowledge atoms, the source basis of the constraint conditions corresponding to the knowledge atoms, and the risk level of the constraint conditions corresponding to the knowledge atoms.

[0108] In one specific implementation, an electronic device deploys a knowledge graph generation model. This model automatically constructs a four-layer ontology structure—knowledge atom, clause type, source basis, and risk level—based on the constraints and classification information corresponding to the knowledge atoms of each constraint. The model can extract entities from the contract verification basis, obtaining multiple entity nodes. Based on the verification rules of the knowledge atoms corresponding to each constraint, the model obtains the association rules between entity nodes. Therefore, based on these association rules, the model can obtain the relationship edges between any two entity nodes and output a first knowledge graph.

[0109] In this embodiment, by breaking down the contract verification criteria into indivisible, verifiable knowledge atoms, a precise foundation for knowledge referencing can be established. Establishing relationships between entities based on the verification rules of these knowledge atoms improves the accuracy of the first knowledge graph's associations, thereby enhancing the accuracy of knowledge graph referencing and reducing the problems of low relevance and incorrect associations in knowledge graph referencing. Furthermore, creating the first knowledge topology via electronic devices eliminates the need for collaboration between professionals and technicians, improving the efficiency of first knowledge graph construction and reducing its cost.

[0110] The method for obtaining knowledge atoms in step S301 above will be described in detail below.

[0111] S3011. The electronic device decomposes the various constraints corresponding to the contract content in the contract verification basis and obtains the initial knowledge atoms corresponding to each constraint.

[0112] Initial knowledge atoms are used to represent knowledge atoms initially extracted from contract verification bases that have not undergone data verification. They have basic guiding value for verifying contract content, but may have issues such as conflicts with existing agreements, disconnect from industry standards, unclear expression, or logical errors. Therefore, data verification of initial knowledge atoms is necessary to improve their accuracy, timeliness, and compliance.

[0113] S3012. The electronic device performs data verification on multiple initial knowledge atoms and uses the initial knowledge atoms that pass the data verification as knowledge atoms.

[0114] In one feasible implementation, data verification includes integrity verification. The electronic device performs integrity verification on the initial knowledge atoms, and uses the initial knowledge atoms that pass the integrity verification as the knowledge atoms.

[0115] Integrity verification refers to verifying the information of the initial knowledge atom, checking whether the initial knowledge atom contains the content required for the basic guiding value of independent contract content verification, avoiding the absence or omission of some content, and ensuring the integrity of the knowledge atom.

[0116] In one specific implementation, electronic devices can verify the integrity and accuracy of knowledge atoms by invoking at least two base models.

[0117] In another feasible implementation, data validation includes normative validation. The electronic device performs normative validation on the initial knowledge atoms based on an expert rule base, and uses the initial knowledge atoms that pass the normative validation as the knowledge atoms.

[0118] The expert rule base includes agreed-upon clauses and industry standards.

[0119] Normative verification refers to verifying the professionalism of initial knowledge atoms and determining whether the initial knowledge atoms meet the agreed terms, industry rules, etc.

[0120] In one specific implementation, data verification can involve introducing a dual-source verification mechanism: the first source is integrity verification, and the second source is normative verification. The electronic device performs integrity verification on the initial knowledge atoms, and also performs normative verification on the initial knowledge atoms based on an expert rule base. The initial knowledge atoms that pass both integrity and normative verifications are then considered as the final knowledge atoms.

[0121] In another specific implementation, data verification can be achieved through a large model and blockchain-based evidence verification. The electronic device is equipped with a large model and stores a blockchain containing the basis for contract verification. The large model verifies the initial knowledge atoms by retrieving on-chain data from the blockchain and uses the verified initial knowledge atoms as the knowledge atoms.

[0122] In this embodiment, by performing integrity and standardization checks on the initial knowledge atoms, errors, omissions, contradictions, and unprofessional content in the obtained knowledge atoms are avoided, ensuring that each knowledge atom is a complete, legal, and reliable verification basis, thereby improving the integrity, accuracy, and professionalism of the knowledge atoms.

[0123] The following describes one method for constructing the first knowledge graph in step S302 above.

[0124] In some embodiments, the electronic device stores knowledge atoms in a knowledge atom database and establishes type indexes corresponding to different types of knowledge atoms based on their types. Thus, the electronic device can retrieve knowledge atoms corresponding to multiple constraints from the knowledge atom database based on the type indexes corresponding to the knowledge atoms of multiple constraints, and construct a first knowledge graph.

[0125] The types of knowledge atoms include source basis, clause compliance, risk identification, and logic verification.

[0126] An index is used to represent a unique identifier corresponding to a knowledge atom type. That is, an electronic device can obtain the knowledge atom of the knowledge atom type corresponding to the index based on an index.

[0127] In this embodiment, by establishing type indexes corresponding to different types of knowledge atoms, it is more convenient for electronic devices to call knowledge atoms, thereby improving the efficiency of electronic devices in constructing the first knowledge graph.

[0128] In some embodiments, the electronic device can acquire contract verification basis based on a preset acquisition cycle. If new contract verification basis is acquired, the electronic device can update the first knowledge graph based on the new contract verification basis to obtain the updated first knowledge graph.

[0129] Electronic devices can collect updated contract verification data from the official websites corresponding to the contract verification clauses and the official websites of various industry departments.

[0130] For example, the preset collection period can be 1 day, 1 week, 1 month, 6 months, 1 year, etc.

[0131] For contract verification criteria that are updated frequently, a shorter preset collection period can be used, for example, collecting the contract verification criteria every week. For contract verification criteria that are not updated frequently, a longer preset collection period can be used, for example, collecting the contract verification criteria every year.

[0132] In one specific implementation, an electronic device is equipped with a knowledge graph generation model. Based on a preset collection period, the knowledge graph generation model collects the contract verification basis from the official websites corresponding to the contract verification clauses and the official websites corresponding to various industry departments. If updated contract verification data is collected, the knowledge graph generation model updates the first knowledge graph through knowledge atom matching and conflict monitoring algorithms.

[0133] In this embodiment, the electronic device periodically collects the updated contract verification basis, deletes invalid content, and adds updated content to avoid the impact of invalid content on contract content verification, thereby improving the timeliness and accuracy of contract content verification. The electronic device enables the updating of the first knowledge graph, eliminating the need for manual updating of entity relationships in the first knowledge graph, thus improving the efficiency of updating the first knowledge graph and reducing the cost of updating the first knowledge graph.

[0134] In other embodiments, the electronic device can monitor the official website corresponding to the contract verification agreement and the official website corresponding to each industry department. When the electronic device detects an updated contract verification basis, it updates the first knowledge graph to obtain the updated first knowledge graph.

[0135] In this embodiment, by monitoring the official websites that update the contract verification basis, the first knowledge graph is updated when the updated contract verification basis is detected. Compared with timed collection, invalid collection is avoided and the timeliness of collection is improved. Furthermore, the timeliness of updating the first knowledge graph is improved.

[0136] The following provides a detailed description of a method for constructing a second knowledge graph based on the target first knowledge graph in step S202 above.

[0137] S2021. Electronic devices extract the target entity from the target contract.

[0138] The target entity includes at least subject-type variables, target-type variables, performance-type variables, or liability-type variables.

[0139] Subject-type variables refer to the actors, responsible parties, and participating parties in the target contract, as well as the characteristics and attributes of each party. For example, subject-type variables include the counterparty's name, unified social credit code, qualification level, etc.

[0140] The subject matter refers to the core transaction object of the target contract. Subject matter class variables are used to characterize the specific characteristics of the core transaction object of the target contract. For example, subject matter class variables include the name of the subject matter, the quantity of the subject matter, the quality standards of the subject matter, and the form of the subject matter.

[0141] Performance-related variables are used to characterize the rights, obligations, operational standards, time boundaries, and behavioral boundaries of both parties to a target contract during the performance phase. For example, performance-related variables include performance period, performance location, payment amount, payment method, payment time, and acceptance period.

[0142] Liability-related variables are used to characterize the attribution of responsibility, the party bearing responsibility, the rules for accountability, and the rules for exemption from liability for the parties signing the target contract at each stage of contract formation, performance, and termination. For example, liability-related variables include the responsible party, the type of responsibility, the conditions for triggering breach of contract, the amount of compensation, and the dispute resolution method.

[0143] The following describes various methods for extracting the target entity of a target contract using electronic devices.

[0144] In one feasible implementation, the electronic device is equipped with a pre-defined field extraction model. The device inputs the target contract into the pre-defined field extraction model for field extraction to obtain the target entity.

[0145] In another feasible implementation, electronic devices can extract the target entity from the target contract using a regular expression matching algorithm.

[0146] Optionally, the pre-defined field extraction model deployed in the electronic device can use a regular expression matching algorithm to extract the target entity of the target contract.

[0147] In another feasible implementation, the electronic device is equipped with an entity recognition model. The electronic device fine-tunes the entity recognition model based on the contract verification criteria of the domain corresponding to the target contract. The target contract is then input into the fine-tuned entity recognition model. The target entity in the target contract is extracted through entity linking technology in the entity recognition model, thereby obtaining the target entity in the target contract.

[0148] S2022, Electronic devices acquire the target first knowledge graph that matches the contract type of the target contract.

[0149] In some embodiments, after acquiring a target contract, the electronic device can obtain contract verification criteria corresponding to the contract type and industry attributes of the target contract. The electronic device can then construct a target first knowledge graph that matches the contract type of the target contract based on these verification criteria.

[0150] In other embodiments, the electronic device may store a first knowledge graph corresponding to various contract types and industry attributes. After acquiring a target contract, the electronic device can directly call the target first knowledge graph that matches the contract type of the target contract based on the contract type and industry attribute of the target contract.

[0151] S2023. The electronic device extracts entity nodes and relation edges that match the target entity from the first knowledge graph and obtains the second knowledge graph corresponding to the target contract.

[0152] The first knowledge graph includes multiple entity nodes and the correspondence between entity nodes. Each entity node includes at least the constraint conditions corresponding to each knowledge atom, the constraint objects of the constraint conditions corresponding to the knowledge atom, the source basis of the constraint conditions corresponding to the knowledge atom, and the risk level of the constraint conditions corresponding to the knowledge atom.

[0153] Electronic devices find entity nodes that match the target entity, extract the entity nodes and relation edges that match the target entity from the first knowledge graph, and establish the association between the target entity, the entity nodes that match the target entity, and the knowledge atoms corresponding to the entity nodes. In this way, the source basis of the target entity, the knowledge atoms corresponding to the target entity, the constraints corresponding to the knowledge atoms, and the risk level of the constraints corresponding to the target knowledge atoms can be obtained, that is, the second knowledge graph is obtained.

[0154] In some embodiments, the first knowledge graph can have a three-level structure, including knowledge atoms, the source basis of the constraints corresponding to the knowledge atoms, and the risk level of the constraints corresponding to the knowledge atoms. Each knowledge atom includes an identifier, the corresponding constraint, the applicable scenario of the constraint, and the verification rules for the constraint. The electronic device constructs a second knowledge graph of the target contract based on this first knowledge graph. The second knowledge graph can have a four-level structure, including a target entity, target knowledge atoms corresponding to the target entity, the source basis of the target constraints corresponding to the target knowledge atoms, and the risk level of the target constraints corresponding to the target knowledge atoms.

[0155] In this embodiment, a second knowledge graph corresponding to the target contract is constructed using a first knowledge graph corresponding to the contract type of the target contract, making the contract content of the target contract reflected in the second knowledge graph more accurate; based on the more accurate second knowledge graph, the verification results obtained subsequently based on the second knowledge graph are more accurate.

[0156] In some embodiments, the electronic device performs logical and integrity checks on the target entity according to the corresponding constraints, and marks the target entity that fails the logical check or the integrity check as a target entity to be verified. The verification result includes the target entity to be verified.

[0157] Logical verification refers to checking whether the target entity in the target contract conforms to objective laws. For example, the performance period cannot be earlier than the contract signing date, or the names of the two parties signing the target contract are inconsistent.

[0158] Integrity verification refers to verifying the integrity of the target entity in the target contract. For example, the target contract may not specify the payment time or payment amount.

[0159] In this embodiment, by performing logical and integrity checks on the target entity, it is possible to avoid the existence of target entities that are illogical, unreasonable, or missing. By marking the target entities that fail the logical or integrity checks in the check results, a prompt can be provided, thereby improving the practicality of the check results.

[0160] The following is a detailed description of one method by which the electronic device obtains the verification result of the target contract based on the second knowledge graph in step S203 above.

[0161] In some embodiments, the electronic device can retrieve a target verification template corresponding to the contract type and domain attributes from a preset verification template library. The electronic device can then populate the target verification template based on the second knowledge graph to obtain the verification result of the target contract.

[0162] When the verification result is a verification opinion letter, the electronic device includes a verification opinion letter template library. The electronic device can obtain the target verification opinion letter template corresponding to the contract type and domain attribute from the verification opinion letter template library based on the contract type and domain attribute. The target contract review opinion letter may include the basic information of the target contract, the verification result of the target contract content, the modification opinions of the target contract, and the verification conclusion opinions, etc.

[0163] The basic information of the target contract may include subject-type variables and object-type variables. The content verification results of the target contract may include content risk items, the source basis corresponding to the content risk items, and the target entity corresponding to the content risk items.

[0164] The proposed modifications to the target contract can be generated based on the verification rules of the knowledge atom. For example, if the verification rule of the knowledge atom is that the deposit amount is 20% of the total contract amount, the corresponding modification suggestion would be to adjust the deposit amount to no more than XX million yuan.

[0165] Verification conclusions can be tailored to the severity of the content risk items. For example, if the content risk items in the verification results for the target contract are of high severity, the verification conclusion could be that the contract needs thorough verification before signing; if the content risk items in the verification results for the target contract are of low severity, the verification conclusion could be that the contract should be verified before signing.

[0166] In one feasible implementation, an electronic device is equipped with a verification template generation model. By inputting a basic verification template and historical verification results into this model, it can output verification templates corresponding to different contract types and domain attributes. The model can also receive feedback from target users corresponding to the target contract and optimize the verification templates for the target contract's contract type and domain attributes based on this feedback.

[0167] Electronic devices can store the verification templates generated by the verification template generation model into a preset verification template library. When the electronic device obtains a target contract of the same contract type and domain attribute for the next time, it can directly call the generated verification template without having to generate a verification template of the same contract type and domain attribute again.

[0168] In this embodiment, the electronic device fills the target verification template with the results of contract content verification based on the second knowledge graph, eliminating the need for manual integration of verification results and improving the efficiency of obtaining verification results. By filling the target verification template with the results of contract content verification, the format of the verification results is unified, and fragmented verification marks are avoided, making it easier for users to continue viewing and improving the user experience.

[0169] In other embodiments, the electronic device is equipped with a verification result generation model that employs prompt engineering-driven generation, enabling the model to directly generate compliant verification results based on refined prompts. For example, the prompts may include information such as contract type, risk items, and source basis.

[0170] In some embodiments, the electronic device can send the verification result of the target contract to the target user corresponding to the target contract through a preset method. The electronic device can then obtain feedback data from the target user, update the target verification template corresponding to the target user's verification result based on this feedback data, store the updated target verification template in a preset verification template library, and update the preset verification template library. Based on the contract type and domain attributes corresponding to the target contract, the electronic device retrieves the target verification template corresponding to the contract type and domain attributes from the updated preset verification template library.

[0171] For example, the preset method could be that the electronic device stores verification results in multiple formats, and the target user corresponding to the target contract can download them from the electronic device to the target user's terminal device.

[0172] Updating the target validation template can involve adjusting its content structure, adding or deleting content, and so on.

[0173] In other embodiments, the electronic device can send the verification result of the target contract to the target user corresponding to the target contract through a preset method, so that the electronic device can obtain the feedback data of the target user, update the knowledge atoms based on the feedback data, obtain the updated knowledge atoms, and enable the electronic device to construct the first knowledge graph based on the updated knowledge atoms corresponding to each constraint condition.

[0174] Updating a knowledge atom can be done by updating the verification rules for that knowledge atom.

[0175] In some embodiments, the electronic device can send the verification result of the target contract to the target user corresponding to the target contract through a preset method. The electronic device can then obtain the feedback data of the target user. Based on the feedback data, the parameters of the regular expression matching algorithm are updated to obtain the updated regular expression matching algorithm. The electronic device can then extract the target entity from the target contract through the updated regular expression matching algorithm.

[0176] In this embodiment, the electronic device updates the knowledge atoms, regular expression matching algorithm, and target verification template based on the target user's feedback data on the verification results, which can improve the accuracy of the electronic device in verifying the contract content.

[0177] In some embodiments, the electronic device can adjust the content structure of the verification result corresponding to the target contract based on the risk levels in the second knowledge graph. For example, if the second knowledge graph includes three high-risk levels and two medium-risk levels, the high-risk level content can be highlighted by moving the content of the high-risk levels before the content of the medium-risk levels.

[0178] In this embodiment, by adjusting the content structure of the verification results and highlighting high-risk content, the practicality of the output verification results can be improved, thereby enhancing the user experience. For the verification results of the target contract, the electronic device can output different forms of verification results based on different conditions.

[0179] In one feasible implementation, the electronic device can output verification results in different formats based on the complexity of the second knowledge graph. For a more complex second knowledge graph, the verification results obtained by the electronic device may include a contract review opinion and the second knowledge graph, or the verification results obtained by the electronic device may include a contract review annotation draft and the second knowledge graph. For a simpler second knowledge graph, the verification results obtained by the electronic device may be a contract verification checklist or the second knowledge graph.

[0180] Based on the complexity of the second knowledge graph, an appropriate format for representing the verification results is selected so that the verification results can better reflect all the content in the second knowledge graph; by outputting verification results in one or more formats, the output verification results are made clearer and the user experience can be improved.

[0181] In another feasible implementation, the electronic device can output different verification results based on the contract type of the target contract. For domain-specific contract types, such as transportation contracts, trust contracts, factoring contracts, medical service contracts, and physical examination service contracts, the verification results obtained by the electronic device can include a contract review opinion and a second knowledge graph, a contract review annotation draft and a second knowledge graph, or a contract verification checklist and a second knowledge graph. For general contract types, the verification results obtained by the electronic device can be at least one of a contract review opinion, a contract review annotation draft, a contract verification checklist, and a second knowledge graph.

[0182] Matching different verification result formats to different contract types can make the output verification results clearer and improve the user experience.

[0183] This application provides a contract content verification system, which includes a knowledge graph generation module and a verification result generation module.

[0184] The knowledge graph generation module is used to construct a second knowledge graph corresponding to the target contract based on the first target knowledge graph that matches the contract type of the target contract.

[0185] The verification result generation module is used to obtain the verification result of the target contract based on the second knowledge graph. The verification result includes at least the content risk items of the target contract.

[0186] In some embodiments, the contract content verification system further includes a knowledge atom construction module and a variable extraction module.

[0187] Figure 4 This is a schematic diagram of the structure of a contract content verification system provided in the embodiments of this application.

[0188] like Figure 4 As shown, the contract content verification system 400 includes an input layer 401, a processing layer 402, and an output layer 403. The processing layer 402 includes a knowledge atom construction module 4021, a knowledge graph generation module 4022, a variable extraction module 4023, and a verification result generation module 4024.

[0189] Input layer 401 is used to obtain the target contract to be verified and the basis for contract verification.

[0190] Processing layer 402 is used to verify the content of the target contract.

[0191] The knowledge atom construction module 4021 is used to decompose the constraints corresponding to the contract content in the contract verification basis and obtain the knowledge atoms corresponding to each constraint.

[0192] In some embodiments, the knowledge atom construction module 4021 is also used to perform data verification on the knowledge atoms, including integrity verification and normative verification based on the expert rule base.

[0193] In other embodiments, the knowledge atom construction module 4021 is also used to store knowledge atoms in a knowledge atom library and to establish type indexes corresponding to different types of knowledge atoms based on the type of knowledge atoms.

[0194] The knowledge graph generation module 4022 is used to extract entities from the contract verification basis, obtain multiple entity nodes, obtain the relationship edges between pairs of entity nodes based on the association rules between entity nodes, and construct the first knowledge graph.

[0195] The knowledge graph generation module 4022 is also used to construct a second knowledge graph corresponding to the target contract based on the target first knowledge graph that matches the contract type of the target contract.

[0196] In some embodiments, the knowledge graph generation module includes a first knowledge graph generation module and a second knowledge graph generation module.

[0197] The first knowledge graph generation module is used to extract entities from the contract verification basis, obtain multiple entity nodes, obtain the relationship edges between pairs of entity nodes based on the association rules between entity nodes, and construct the first knowledge graph.

[0198] The first knowledge graph generation module is also used to collect contract verification criteria based on a preset collection period. If updated contract verification criteria are collected, the electronic device updates the first knowledge graph based on the updated contract verification criteria to obtain the updated first knowledge graph.

[0199] In one specific implementation, the first knowledge graph generation module is equipped with a first knowledge graph generation model, which is used to implement the functions of the first knowledge graph module.

[0200] The second knowledge graph generation module is used to construct a second knowledge graph corresponding to the target contract based on the target first knowledge graph that matches the contract type of the target contract.

[0201] The second knowledge graph generation module is also used to store the second knowledge graph in a structured form and provide a visualization interface. This visualization interface is used to show users the relationship between the entity nodes corresponding to the target entity of the target contract and the risk transmission path, thereby improving the user experience.

[0202] In one specific implementation, a second knowledge graph generation model is deployed in the second knowledge graph generation module. This second knowledge graph generation model is used to implement the functions of the aforementioned second knowledge graph generation module.

[0203] The variable extraction module 4023 is used to extract the target entity from the target contract.

[0204] In some embodiments, the variable extraction module 4023 is further configured to input the target contract into a preset large model, extract fields through the regular expression matching algorithm of the preset large model, and obtain the target entity of the target contract.

[0205] In other embodiments, the variable extraction module 4023 is further configured to perform logical and integrity checks on the target entities corresponding to the constraints, and mark target entities that fail the logical or integrity checks as target entities to be verified. The verification results include the target entities to be verified.

[0206] The verification result generation module 4024 is used to obtain the verification result of the target contract based on the second knowledge graph. The verification result includes at least the content risk items of the target contract.

[0207] In some embodiments, the verification result generation module 4024 is further configured to obtain a target verification template corresponding to the contract type and domain attribute from a preset verification template library based on the contract type and domain attribute corresponding to the target contract, fill the target verification template based on the second knowledge graph, and obtain the verification result of the target contract.

[0208] In other embodiments, the verification result generation module 4024 is also used to optimize the generated verification result to ensure the professionalism, logic and accuracy of the verification result. The verification result also supports user-defined supplementary remarks.

[0209] In some embodiments, the processing layer 402 of the contract content verification system 400 further includes an output module 4025, which is used to send the verification result of the target contract to the target user corresponding to the target contract in a preset manner.

[0210] The output module 4025 is also used to obtain feedback data from the target user, update the knowledge atom based on the feedback data, obtain the updated knowledge atom, update the regular expression matching algorithm based on the feedback data, obtain the updated regular expression matching algorithm, and update the target verification template corresponding to the verification result of the target user based on the feedback data, obtain the updated target verification template.

[0211] Output layer 403 is used to send the verification result corresponding to the target contract to the target user corresponding to the target contract.

[0212] Figure 5 This is a schematic diagram of a contract content verification device provided in an embodiment of this application. Figure 5 The contract content verification device 500 shown includes a processing module 501, a communication module 502, and a storage module 503.

[0213] The processing module 501 may be a central processing unit (CPU), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. The processor may include an application processor and a baseband processor. It may implement or execute the various exemplary logic blocks, modules, and circuits described in connection with the disclosure of this application. The processor may also be a combination that implements computational functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0214] For example, the processing module 501 can be as follows: Figure 1 The processor 102 shown; the communication module 502 can be as follows: Figure 1 The communication interface 104 shown; the storage module 503 can be as follows: Figure 1 The internal memory 103 shown.

[0215] Through the above description of the embodiments, those skilled in the art will clearly understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device, and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0216] This application provides a computer program product containing instructions that, when run on a computer, cause the computer to execute the contract content verification method described in the above method embodiments.

[0217] This application also provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the contract content verification method in the method flow shown in the above method embodiments.

[0218] The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), a register, a hard disk, an optical fiber, a compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof, or any other form of computer-readable storage medium known in the art. An exemplary storage medium is coupled to a processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium may also be a component of the processor. The processor and the storage medium may reside in an application-specific integrated circuit (ASIC). In the embodiments of this application, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0219] Since the contract content verification device, computer-readable storage medium, and computer program product in the embodiments of the present invention can be applied to the above methods, the technical effects they can achieve can also be referred to the above method embodiments. The embodiments of the present invention will not be described again here.

[0220] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0221] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0222] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0223] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for verifying contract content, characterized in that, The invention is applied to an electronic device, which stores a first knowledge graph corresponding to at least one type of contract in at least one domain. The first knowledge graph is used to characterize the relationship between the contract content corresponding to a contract type, the source basis of the contract content, and the risk level of the contract content. The method includes: The electronic device acquires the target contract; The electronic device constructs a second knowledge graph corresponding to the target contract based on a first knowledge graph that matches the contract type of the target contract; the second knowledge graph is used to characterize the relationship between the target contract content and the source basis of the target contract content, and the risk level of the target contract content. The electronic device obtains the verification result of the target contract based on the second knowledge graph; the verification result includes at least the content risk item of the target contract.

2. The method according to claim 1, characterized in that, The method further includes: The electronic device decomposes the constraints corresponding to the contract content in the contract verification basis to obtain knowledge atoms corresponding to each constraint; the knowledge atom includes an identifier, the corresponding constraint, the applicable scenario of the corresponding constraint, and the verification rule of the corresponding constraint. The electronic device constructs the first knowledge graph based on the knowledge atoms corresponding to each constraint condition.

3. The method according to claim 2, characterized in that, The electronic device constructs the first knowledge graph based on the knowledge atoms corresponding to each constraint condition, including: The electronic device extracts entities from the contract verification basis based on each of the knowledge atoms to obtain multiple entity nodes; each entity node includes at least the constraint condition corresponding to each knowledge atom, the constraint object of the constraint condition corresponding to the knowledge atom, the source basis of the constraint condition corresponding to the knowledge atom, and the risk level of the constraint condition corresponding to the knowledge atom. The electronic device obtains the relationship edges between pairs of entity nodes based on the association rules between the entity nodes, and constructs the first knowledge graph.

4. The method according to claim 2 or 3, characterized in that, The method further includes: The electronic device acquires contract verification criteria based on a preset acquisition cycle. If new contract verification criteria are acquired, the electronic device updates the first knowledge graph based on the new contract verification criteria to obtain an updated first knowledge graph.

5. The method according to any one of claims 2-4, characterized in that, The electronic device decomposes the constraints corresponding to the contract content in the contract verification basis to obtain knowledge atoms corresponding to each constraint, including: The electronic device decomposes the various constraints corresponding to the contract content in the contract verification basis and obtains the initial knowledge atoms corresponding to each constraint. The electronic device performs data verification on the plurality of initial knowledge atoms, and uses the initial knowledge atoms that pass the data verification as the knowledge atoms.

6. The method according to claim 5, characterized in that, The electronic device performs data verification on the plurality of initial knowledge atoms, and uses the initial knowledge atoms that pass the data verification as the knowledge atoms, including: The electronic device performs integrity verification on the plurality of initial knowledge atoms; The electronic device performs normative verification on the initial knowledge atoms based on an expert rule base; The electronic device uses the initial knowledge atoms obtained through integrity verification and normativity verification as the knowledge atoms.

7. The method according to any one of claims 2-6, characterized in that, The method further includes: The electronic device stores the knowledge atoms in a knowledge atom database and establishes type indexes corresponding to different types of knowledge atoms based on the type of the knowledge atoms. The electronic device constructs the first knowledge graph based on knowledge atoms corresponding to multiple constraints, including: The electronic device retrieves the knowledge atoms corresponding to the multiple constraints from the knowledge atom database based on the type index of the knowledge atoms corresponding to the multiple constraints, and constructs the first knowledge graph.

8. The method according to any one of claims 1-7, characterized in that, The electronic device constructs a second knowledge graph corresponding to the target contract based on a first knowledge graph that matches the contract type of the target contract, including: The electronic device extracts the target entity from the target contract; the target entity includes at least subject-type variables, subject-type variables, performance-type variables, or liability-type variables; The electronic device acquires a target first knowledge graph that matches the contract type of the target contract; The electronic device extracts entity nodes and relation edges that match the target entity from the first knowledge graph, and obtains a second knowledge graph corresponding to the target contract.

9. The method according to claim 8, characterized in that, The electronic device extracts the target entity from the target contract, including: The electronic device inputs the target contract into a preset field extraction model for field extraction to obtain the target entity; Alternatively, the electronic device may extract the target entity from the target contract using a regular expression matching algorithm.

10. The method according to claim 8 or 9, characterized in that, After the electronic device extracts the target entity from the target contract, the method further includes: The electronic device performs logical and integrity checks on the target entity according to the corresponding constraints, and marks the target entity that fails the logical check or the integrity check as a target entity to be verified; the verification result includes the target entity to be verified.

11. The method according to any one of claims 8-10, characterized in that, The first knowledge graph has a three-level structure, which includes knowledge atoms, the source basis of the constraints corresponding to the knowledge atoms, and the risk level of the constraints corresponding to the knowledge atoms. The second knowledge graph has a four-level structure, which includes a target entity, a target knowledge atom corresponding to the target entity, the source basis of the target constraint conditions corresponding to the target knowledge atom, and the risk level of the target constraint conditions corresponding to the target knowledge atom.

12. The method according to any one of claims 1-11, characterized in that, The electronic device obtains the verification result of the target contract based on the second knowledge graph, including: The electronic device obtains a target verification template corresponding to the contract type and domain attribute from a preset verification template library based on the contract type and domain attribute corresponding to the target contract. The electronic device fills the target verification template based on the second knowledge graph to obtain the verification result of the target contract.

13. The method according to claim 12, characterized in that, The method further includes: The electronic device sends the verification result of the target contract to the target user corresponding to the target contract in a preset manner; The electronic device acquires feedback data from the target user; Based on the feedback data, the electronic device updates the target verification template corresponding to the verification result of the target user, stores the updated target verification template in a preset verification template library, and updates the preset verification template library. The electronic device, based on the contract type and domain attribute corresponding to the target contract, retrieves a target verification template corresponding to the contract type and domain attribute from a preset verification template library, including: The electronic device retrieves the target verification template corresponding to the contract type and the domain attribute from the updated preset verification template library based on the contract type and domain attribute corresponding to the target contract.

14. The method according to any one of claims 2-12, characterized in that, The method further includes: The electronic device sends the verification result of the target contract to the target user corresponding to the target contract in a preset manner; The electronic device acquires feedback data from the target user; The electronic device updates the knowledge atom based on the feedback data and obtains the updated knowledge atom. The electronic device constructs the first knowledge graph based on the knowledge atoms corresponding to each constraint condition, including: The electronic device constructs the first knowledge graph based on the updated knowledge atoms corresponding to each constraint condition.

15. The method according to any one of claims 9-12, characterized in that, The method further includes: The electronic device sends the verification result of the target contract to the target user corresponding to the target contract in a preset manner; The electronic device acquires feedback data from the target user; The electronic device updates the parameters of the regular expression matching algorithm based on the feedback data to obtain the updated regular expression matching algorithm. The electronic device extracts the target entity from the target contract using a regular expression matching algorithm, including: The electronic device extracts the target entity from the target contract using an updated regular expression matching algorithm.

16. A contract review system, characterized in that, The contract review system includes a knowledge graph generation module and a verification result generation module; The knowledge graph generation module is used to construct a second knowledge graph corresponding to the target contract based on a first target knowledge graph that matches the contract type of the target contract. The verification result generation module is used to obtain the verification result of the target contract based on the second knowledge graph; The verification results include at least the content risk items of the target contract.

17. An electronic device, characterized in that, The electronic device includes: a display, a processor, a communication interface, and a memory, wherein the memory stores at least one computer program, and the at least one computer program is loaded and executed by the processor to implement the contract content verification method as described in any one of claims 1 to 15.

18. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one computer program, which is loaded and executed by a processor to implement the contract content verification method as described in any one of claims 1 to 15.

19. A computer program product, characterized in that, The computer program product includes a computer program or instructions that, when executed by a processor, implement the contract content verification method as described in any one of claims 1 to 15.

Images