Contract term conflict detection method and device, equipment, storage medium and program product

Abstract

The invention relates to a contract term conflict detection method and device, equipment, a storage medium and a program product. The method comprises the steps of identifying a key entity in a to-be-audited contract text, segmenting the to-be-audited contract text according to chapters and clause numbers of the to-be-audited contract text to obtain a first clause object set of the to-be-audited contract text, obtaining a semantic vector and obligational condition relationship pair corresponding to the first clause object set, and obtaining a semantic vector and obligational condition relationship pair corresponding to the first clause object set; according to the semantic vector corresponding to the first term object set, the obligational condition relationship pair and the key entity, performing conflict detection on terms of the contract text to be audited by using the conflict detection model to obtain a conflict detection result, thereby realizing identification of logic contradictions and semantic conflicts among the contract terms and potential problems in a dynamic scene, and improving the success rate of contract verification. The efficiency of contract review can be improved.

Description

Technical Field

[0001] The present application relates to the field of data analysis technology, and in particular to a contract clause conflict detection method, apparatus, device, storage medium, and program product. Background Art

[0002] In the field of platform-based e-commerce, the application of electronic contracts has become relatively mature. At present, the online push of standard structured contracts, online signing, and automatic application of structured terms data have basically been realized.

[0003] At present, the current electronic contract functions still have shortcomings, especially in the automated review of merchant contracts, and a large amount of manual review and confirmation of specific contract terms is still required.

[0004] However, manual review relies on the experience of legal experts and has the problem of low efficiency. Summary of the Invention

[0005] Based on this, it is necessary to provide a contract clause conflict detection method, device, equipment, storage medium and program product that can improve the efficiency of contract clause conflict detection in response to the above technical problems.

[0006] In a first aspect, the present application provides a method for detecting a conflict in contract terms, comprising:

[0007] Identify key entities in the contract to be reviewed; key entities include at least the contract subject, amount, and time point;

[0008] Split the contract text to be reviewed according to its chapter and clause numbers, and obtain a set of first clause objects of the contract text to be reviewed;

[0009] Obtain the semantic vector and obligation condition relationship pair corresponding to the first clause object set;

[0010] According to the semantic vectors, obligation-condition relationship pairs and key entities corresponding to the first clause object set, the conflict detection model is used to perform conflict detection on the clauses of the contract text to be reviewed, and the conflict detection results are obtained.

[0011] In one embodiment, the method further comprises:

[0012] Obtain historical contract texts, legal provisions, and precedents on contract disputes;

[0013] Split the historical contract text according to its chapter and clause numbers to obtain a set of second clause objects of the historical contract text;

[0014] The second clause object set, legal provisions, and contract dispute precedents are treated as nodes, and the relationship between each node is obtained;

[0015] Construct an initial knowledge graph based on each node and the relationship between them;

[0016] An initial conflict detection model is trained based on the initial knowledge graph to obtain a conflict detection model.

[0017] In one embodiment, the method further comprises:

[0018] Obtain feedback on conflict detection results;

[0019] Update the initial knowledge graph according to the feedback information to obtain a new knowledge graph;

[0020] The conflict detection model is trained based on the new knowledge graph to optimize the conflict detection model.

[0021] In one embodiment, obtaining the relationship between nodes includes:

[0022] The conflict type and association strength weight between each node are obtained, and the relationship between each node is determined according to the conflict type and association strength weight between each node.

[0023] In one embodiment, based on the semantic vectors, obligation-condition relationship pairs, and key entities corresponding to the first clause object set, a conflict detection model is used to perform conflict detection on the clauses of the contract text to be reviewed, and conflict detection results are obtained, including:

[0024] Determine the edge features between the clause object set groups based on the semantic vectors, obligation-condition relationship pairs, and key entities corresponding to each clause object set group; the clause object set group includes any two first clause object sets;

[0025] For each clause object set group, based on the edge features and corresponding semantic vectors between the clause object set groups, a conflict detection model is used to determine the conflict probability between the two first clause object sets in the clause object set group;

[0026] Based on the conflict probability corresponding to each clause object set group, the conflict detection result of the contract text to be reviewed is obtained.

[0027] In one embodiment, obtaining the semantic vector and obligation-condition relationship pair corresponding to the first clause object set includes:

[0028] The BERT model is used to represent the bidirectional encoder of the target transformer to obtain the semantic vector corresponding to the first clause object set, and the first clause object set is parsed based on the dependency syntax tree to obtain the obligation-condition relationship pair.

[0029] In a second aspect, the present application further provides a contract clause conflict detection device, comprising:

[0030] Identification module, used to identify key entities in the contract text to be reviewed; key entities include at least the contract subject, amount, and time point;

[0031] The first segmentation module is used to segment the contract text to be reviewed according to its chapters and clause numbers to obtain a set of first clause objects of the contract text to be reviewed;

[0032] A first acquisition module is used to obtain semantic vectors and obligation-condition relationship pairs corresponding to a first set of clause objects;

[0033] The detection module is used to perform conflict detection on the clauses of the contract text to be reviewed using a conflict detection model based on the semantic vectors, obligation-condition relationship pairs and key entities corresponding to the first clause object set to obtain conflict detection results.

[0034] In a third aspect, the present application further provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0035] Identify key entities in the contract to be reviewed; key entities include at least the contract subject, amount, and time point;

[0036] Split the contract text to be reviewed according to its chapter and clause numbers, and obtain a set of first clause objects of the contract text to be reviewed;

[0037] Obtain the semantic vector and obligation condition relationship pair corresponding to the first clause object set;

[0038] According to the semantic vectors, obligation-condition relationship pairs and key entities corresponding to the first clause object set, the conflict detection model is used to perform conflict detection on the clauses of the contract text to be reviewed, and the conflict detection results are obtained.

[0039] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the following steps:

[0040] Identify key entities in the contract to be reviewed; key entities include at least the contract subject, amount, and time point;

[0041] Split the contract text to be reviewed according to its chapter and clause numbers, and obtain a set of first clause objects of the contract text to be reviewed;

[0042] Obtain the semantic vector and obligation condition relationship pair corresponding to the first clause object set;

[0043] According to the semantic vectors, obligation-condition relationship pairs and key entities corresponding to the first clause object set, the conflict detection model is used to perform conflict detection on the clauses of the contract text to be reviewed, and the conflict detection results are obtained.

[0044] In a fifth aspect, the present application further provides a computer program product having a computer program stored thereon, which, when executed by a processor, implements the following steps:

[0045] Identify key entities in the contract to be reviewed; key entities include at least the contract subject, amount, and time point;

[0046] Split the contract text to be reviewed according to its chapter and clause numbers, and obtain a set of first clause objects of the contract text to be reviewed;

[0047] Obtain the semantic vector and obligation condition relationship pair corresponding to the first clause object set;

[0048] According to the semantic vectors, obligation-condition relationship pairs and key entities corresponding to the first clause object set, the conflict detection model is used to perform conflict detection on the clauses of the contract text to be reviewed, and the conflict detection results are obtained.

[0049] The above-mentioned contract clause conflict detection method, device, equipment, storage medium and program product identify key entities in the contract text to be reviewed, divide the contract text to be reviewed according to the chapter and clause numbers of the contract text to be reviewed, obtain the first clause object set of the contract text to be reviewed, obtain the semantic vector and obligation-condition relationship pair corresponding to the first clause object set, and use the conflict detection model to perform conflict detection on the clauses of the contract text to be reviewed based on the semantic vector, obligation-condition relationship pair and key entities corresponding to the first clause object set to obtain conflict detection results, thereby realizing the identification of logical contradictions, semantic conflicts and potential problems in dynamic scenarios between contract clauses. Since there is no need for manual review of the contract and no reliance on the experience of legal experts, the conflict detection model is used to perform conflict detection on the clauses of the contract text to be reviewed, thereby improving the efficiency of contract review. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0051] Figure 1 This is a flow chart of a method for detecting a conflict in contract terms provided in this embodiment;

[0052] Figure 2 This is a flow chart of a method for obtaining a conflict detection model provided in this embodiment;

[0053] Figure 3 is a flow chart of a method for continuously updating a conflict detection model provided in this embodiment;

[0054] Figure 4 is a flow chart of a method for obtaining conflict detection results provided in this embodiment;

[0055] Figure 5 This is a schematic diagram of the overall architecture of a contract clause conflict detection method provided in this embodiment;

[0056] Figure 6 This is a structural block diagram of a contract clause conflict detection device provided in an embodiment of the present application;

[0057] Figure 7 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0058] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0059] In an exemplary embodiment, Figure 1 As shown, Figure 1 This is a flow chart of a contract clause conflict detection method provided in this embodiment. The method is described by taking the application of the method to a computer device as an example. The method includes the following steps S101 to S104:

[0060] S101, identify key entities in the contract text to be reviewed; key entities include at least the contract subject, amount, and time point.

[0061] A contracting party refers to the party that enjoys rights and bears obligations in a contractual legal relationship. Time points can include information such as the contract's effective date, contract fulfillment period, and deadline. The contract text to be reviewed can be an electronic contract text used on an e-commerce platform.

[0062] S102: Segment the contract text to be reviewed according to its chapters and clause numbers to obtain a first clause object set of the contract text to be reviewed.

[0063] The first clause object set can be represented by C, where C={c1,c2,...,cn}, c1 represents the first clause in C, c2 represents the second clause in C, and cn represents the nth clause in C.

[0064] S103. Obtain the semantic vectors corresponding to the first clause object set and the obligation condition relation pairs.

[0065] The semantic vectors corresponding to the first clause object set can be obtained by using the Bidirectional Encoder Representations from Transformers (BERT) model of the target from the converter, and the obligation condition relation pairs can be obtained by parsing the first clause object set based on the dependency syntactic tree. Here, BERT is the abbreviation of Bidirectional Encoder Representations from Transformers. The target BERT model in this embodiment can be a Legal-BERT model. The Legal-BERT model is a model obtained by pre-training on a large-scale legal corpus based on the BERT model architecture. The Legal-BERT model can output 768-dimensional semantic vectors Vi corresponding to the first clause object set Ci.

[0066] In the extraction of obligation condition relation pairs, the key lies in identifying specific dependency relation patterns related to obligations and conditions. For example, verbs indicating obligations such as "shall" and "must" may have corresponding nominal subjects as the bearers of obligations, as well as some adverbial components to describe the conditions or scopes of obligations. Take the sentence "Party B shall complete the task within 10 days" as an example. Through dependency syntactic analysis, it can be determined that "shall" is the core verb, and "Party B" is the nominal subject of "shall", representing the bearer of the obligation; "within 10 days" is an adverbial modification, indicating the time condition for completing the task, thus forming an obligation condition relation pair like "(Party B, complete the task within 10 days)".

[0067] S104. According to the semantic vectors corresponding to the first clause object set, the obligation condition relation pairs, and the key entities, use the conflict detection model to perform conflict detection on the clauses of the contract text to be reviewed, and obtain the conflict detection result.

[0068] Among them, the conflict detection model can be a Graph Attention Network (GAT) model. The conflict detection model can detect potential conflict paths between clause nodes and obtain the conflict detection result. The conflict detection model can automatically detect logical contradictions, legal compliance conflicts, and semantic ambiguities in the contract text to be reviewed, thereby improving the comprehensiveness of conflict detection and making the conflict detection result more accurate.

[0069] Conflict detection results can include conflict details, such as the location of the conflicting clause, the conflict type, and the legal basis. They can also include suggested corrections, such as adjusting the contract's performance timeline, adding disclaimers, logical adjustments, text rewrites, and compliance supplements. Logical adjustments include, for example, reordering clause execution priorities, text rewrites include generating unambiguous statements based on a template library, and compliance supplements include inserting required legal disclaimers.

[0070] In this embodiment, by identifying key entities in the contract text to be reviewed, the contract text to be reviewed is divided according to the chapters and clause numbers of the contract text to be reviewed, and the first clause object set of the contract text to be reviewed is obtained. The semantic vectors and obligation-condition relationship pairs corresponding to the first clause object set are obtained. According to the semantic vectors, obligation-condition relationship pairs and key entities corresponding to the first clause object set, a conflict detection model is used to perform conflict detection on the clauses of the contract text to be reviewed, and conflict detection results are obtained. This realizes the identification of logical contradictions, semantic conflicts and potential problems in dynamic scenarios between contract clauses. Since there is no need for manual review of the contract and no reliance on the experience of legal experts, the conflict detection model is used to perform conflict detection on the clauses of the contract text to be reviewed, thereby improving the efficiency of contract review.

[0071] In an exemplary embodiment, Figure 2 As shown, Figure 2 FIG. 1 is a flow chart of a method for obtaining a conflict detection model provided in this embodiment, which includes the following steps:

[0072] S201, obtain historical contract texts, legal provisions, and contract dispute precedents.

[0073] S202 , dividing the historical contract text according to its chapter and clause numbers to obtain a second clause object set of the historical contract text.

[0074] S203: The second clause object set, legal provisions, and contract dispute precedents are respectively taken as nodes, and the relationship between the nodes is obtained.

[0075] The conflict type and association strength weight between each node can be obtained, and the relationship between each node can be determined based on the conflict type and association strength weight between each node.

[0076] Examples of conflict types include "time overlap" and "responsibility conflict." "Time overlap" refers to conflicting time requirements across different clauses, potentially preventing simultaneous performance. "Responsibility conflict" refers to conflicting responsibilities across different clauses, making it difficult for the parties to clearly define their obligations. These conflict types can be defined as edge attributes to represent conflict relationships between nodes.

[0077] Based on factors such as the closeness of the logical relationship between nodes and semantic similarity, edge association strength weights are assigned. For example, if a clause is highly semantically and logically related to a legal provision, the edge association strength weight between them will be higher; if the relationship is merely a general reference, the weight will be lower. This approach allows for a more precise description of the degree of connection between nodes in the knowledge graph.

[0078] Edge features can be determined based on the conflict type and association strength weight between nodes. Edges are constructed based on the edge features, and the relationships between nodes can be reflected through the edges. The edge features can include the conflict type and association strength weight between nodes.

[0079] S204: Construct an initial knowledge graph based on each node and the relationship between each node.

[0080] S205: Train an initial conflict detection model based on the initial knowledge graph to obtain a conflict detection model.

[0081] In this embodiment, by obtaining historical contract texts, legal provisions, and contract dispute precedents, the historical contract texts are segmented according to their chapters and clause numbers to obtain a set of second clause objects from the historical contract texts. The set of second clause objects, legal provisions, and contract dispute precedents are then used as nodes. The relationships between the nodes are then obtained. An initial knowledge graph is constructed based on the nodes and their relationships. An initial conflict detection model is trained based on the initial knowledge graph to obtain a conflict detection model, thereby improving the accuracy of the resulting conflict detection model. Furthermore, because the initial knowledge graph integrates legal provisions and contract dispute precedents as external knowledge nodes, the conflict detection model can support compliance verification in multiple legal domains.

[0082] In an exemplary embodiment, Figure 3 As shown, Figure 3 FIG. 4 is a flow chart of a method for continuously updating a conflict detection model provided in this embodiment, which includes the following steps:

[0083] S301: Obtain feedback information regarding conflict detection results.

[0084] The user can provide feedback information regarding the conflict detection results obtained. For example, the feedback information includes that there is no actual time overlap in the clauses but the conflict detection results identify that there is time overlap. That is, when the conflict detection results are inconsistent with the actual situation, the user can provide feedback information regarding the conflict detection results.

[0085] S302: Update the initial knowledge graph according to the feedback information to obtain a new knowledge graph.

[0086] The new knowledge graph is obtained by updating the initial knowledge graph according to the feedback information, thereby updating the knowledge graph. The conflict detection model is trained based on the newly obtained knowledge graph, so that the detection of the obtained conflict detection model is more accurate.

[0087] S303: Training a conflict detection model based on the new knowledge graph to optimize the conflict detection model.

[0088] In this embodiment, by obtaining feedback information on the conflict detection results, the initial knowledge graph is updated according to the feedback information to obtain a new knowledge graph, and the conflict detection model is trained based on the new knowledge graph to optimize the conflict detection model. In the actual application of the conflict detection model, the pre-trained conflict detection model can be optimized in combination with the actual conflict detection results, so that the detection results of the optimized conflict detection model are more accurate.

[0089] In an exemplary embodiment, the above-mentioned acquisition of the relationship between the nodes can be achieved by:

[0090] The conflict type and association strength weight between each node are obtained, and the relationship between each node is determined according to the conflict type and association strength weight between each node.

[0091] In this embodiment, by obtaining the conflict type and association strength weight between each node, and determining the relationship between each node based on the conflict type and association strength weight between each node, the relationship between the nodes can be measured from two aspects: the conflict type and the association strength weight, thereby improving the accuracy and comprehensiveness of the relationship between the nodes.

[0092] In an exemplary embodiment, Figure 4 As shown, Figure 4 FIG. 1 is a flow chart of a method for obtaining conflict detection results provided in this embodiment, which includes the following steps:

[0093] S401, determining edge features between clause object sets according to semantic vectors, obligation-condition relationship pairs and key entities corresponding to each clause object set group; a clause object set group includes any two first clause object sets.

[0094] S402 : For each clause object set group, based on the edge features and corresponding semantic vectors between the clause object set groups, a conflict detection model is used to determine the conflict probability between two first clause object sets in the clause object set group.

[0095] Pconflict(Ci,Cj)=σ(GAT(Vi,Vj, )) Among them, σ is the Sigmoid function, Represents the edge feature between node i corresponding to the i-th first-clause object set and node j corresponding to the j-th first-clause object set. Ci represents the i-th first-clause object set, Cj represents the j-th first-clause object set, Vi represents the semantic vector corresponding to the i-th first-clause object set, and Vj represents the semantic vector corresponding to the j-th first-clause object set. Pconflict(Ci, Cj) represents the conflict probability between the i-th first-clause object set and the j-th first-clause object set.

[0096] The above formula of Pconflict(Ci, Cj) can be used to calculate the conflict probability between two first clause object sets in the clause object set group.

[0097] S403: Obtain conflict detection results of the contract text to be reviewed based on the conflict probabilities corresponding to the set of clause objects.

[0098] In this embodiment, the edge features between the clause object set groups are determined based on the semantic vectors, obligation-condition relationship pairs and key entities corresponding to each clause object set group. For each clause object set group, the conflict probability between the two first clause object sets in the clause object set group is determined based on the edge features and corresponding semantic vectors between the clause object set groups. Based on the conflict probabilities corresponding to each clause object set group, the conflict detection results of the contract text to be reviewed are obtained, thereby improving the efficiency and accuracy of contract review.

[0099] In an exemplary embodiment, the above S103 can be implemented as follows:

[0100] The BERT model is used to represent the bidirectional encoder of the target transformer to obtain the semantic vector corresponding to the first clause object set, and the first clause object set is parsed based on the dependency syntax tree to obtain the obligation-condition relationship pair.

[0101] The target BERT model can be used to obtain the semantic vectors corresponding to the first set of clause objects and to parse the first set of clause objects based on the dependency syntax tree to obtain the obligation-condition relationship pairs. BERT stands for Bidirectional Encoder Representations from Transformers. In this embodiment, the target BERT model can be a legal BERT model, which is based on the BERT model architecture and is pre-trained on a large-scale legal corpus.

[0102] In this embodiment, the target BERT model is used to obtain the semantic vector corresponding to the first clause object set, and the first clause object set is parsed based on the dependency syntax tree to obtain the obligation-condition relationship pair, so as to provide a basis for the subsequent conflict detection of the clauses of the contract text to be reviewed based on the semantic vector corresponding to the first clause object set, the obligation-condition relationship pair and the key entity, and the conflict detection result.

[0103] In an exemplary embodiment, Figure 5 As shown, Figure 5 This is a schematic diagram of the overall architecture of a contract clause conflict detection method provided in this embodiment. The method may include obtaining historical contract texts from various e-commerce platforms based on obtaining external data sources, which may include legal provisions, contract dispute precedents, and other data sources. The historical contract text is segmented according to the chapters and clause numbers of the historical contract text to obtain the second clause object set of the historical contract text, the second clause object set, the legal provisions and the contract dispute precedents are respectively used as nodes, and the relationship between each node is obtained, an initial knowledge graph is constructed according to each node and the relationship between each node, and an initial conflict detection model is trained based on the initial knowledge graph to obtain a conflict detection model; key entities in the contract text to be reviewed are identified, the contract text to be reviewed is segmented according to the chapters and clause numbers of the contract text to be reviewed to obtain the first clause object set of the contract text to be reviewed, the semantic vectors and obligation-condition relationship pairs corresponding to the first clause object set are obtained, and the conflict detection model is used to perform conflict detection on the clauses of the contract text to be reviewed based on the semantic vectors, obligation-condition relationship pairs and key entities corresponding to the first clause object set to obtain conflict detection results; feedback information on the conflict detection results is obtained, the initial knowledge graph is updated according to the feedback information to obtain a new knowledge graph, and the conflict detection model is trained based on the new knowledge graph to achieve incremental learning to optimize the conflict detection model.

[0104] In this embodiment, since the conflict detection model is used to perform conflict detection on the clauses of the contract text to be reviewed, the conflict detection results are obtained without manual review. Therefore, there is no need to rely on the experience of legal experts, and the accuracy and efficiency of contract review can be improved, and the probability of missed or erroneous review of contract clauses can be reduced.

[0105] It should be understood that, although the steps in the flowcharts of the above embodiments are shown in sequence as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be performed in other orders. Moreover, at least a portion of the steps in the flowcharts of the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily to be performed in sequence, but can be performed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0106] Based on the same inventive concept, embodiments of the present application also provide a contract clause conflict detection device for implementing the aforementioned contract clause conflict detection method. The solution provided by this device is similar to the solution described in the aforementioned method. Therefore, the specific limitations of one or more embodiments of the contract clause conflict detection device provided below can be found in the aforementioned limitations of the contract clause conflict detection method and will not be further elaborated here.

[0107] In an exemplary embodiment, Figure 6 As shown, Figure 6 6 is a block diagram of a device for detecting a contract clause conflict provided in an embodiment of the present application. The device 600 includes:

[0108] Identification module 601 is used to identify key entities in the contract text to be reviewed; key entities include at least the contract subject, amount, and time point;

[0109] The first segmentation module 602 is used to segment the contract text to be reviewed according to its chapters and clause numbers to obtain a first clause object set of the contract text to be reviewed;

[0110] A first acquisition module 603 is used to acquire semantic vectors and obligation-condition relationship pairs corresponding to a first set of clause objects;

[0111] The detection module 604 is used to perform conflict detection on the clauses of the contract text to be reviewed using a conflict detection model based on the semantic vectors, obligation-condition relationship pairs and key entities corresponding to the first clause object set to obtain a conflict detection result.

[0112] In one embodiment, the apparatus 600 may further include:

[0113] The second acquisition module is used to obtain historical contract texts, legal provisions, and contract dispute precedents;

[0114] A second segmentation module is used to segment the historical contract text according to its chapter and clause numbers to obtain a set of second clause objects of the historical contract text;

[0115] The third acquisition module is used to treat the second clause object set, legal provisions, and contract dispute precedents as nodes, and obtain the relationship between the nodes;

[0116] The construction module is used to construct the initial knowledge graph based on each node and the relationship between nodes;

[0117] The training module is used to train the initial conflict detection model based on the initial knowledge graph to obtain the conflict detection model.

[0118] In one embodiment, the apparatus 600 further includes:

[0119] The fourth acquisition mode is used to obtain feedback information on the conflict detection result;

[0120] The updating module is used to update the initial knowledge graph according to the feedback information to obtain a new knowledge graph;

[0121] The update module is also used to train the conflict detection model based on the new knowledge graph to optimize the conflict detection model.

[0122] In one embodiment, the third acquisition module is specifically configured to acquire the conflict type and association strength weight between each node, and determine the relationship between each node according to the conflict type and association strength weight between each node.

[0123] In one embodiment, the detection module 604 is specifically used to determine the edge features between the clause object set groups based on the semantic vectors, obligation-condition relationship pairs and key entities corresponding to each clause object set group; the clause object set group includes any two first clause object sets; for each clause object set group, based on the edge features and corresponding semantic vectors between the clause object set groups, a conflict detection model is used to determine the conflict probability between the two first clause object sets in the clause object set group; based on the conflict probability corresponding to each clause object set group, the conflict detection result of the contract text to be reviewed is obtained.

[0124] In one embodiment, the first acquisition module 603 is specifically used to obtain the semantic vector corresponding to the first clause object set using the bidirectional encoder representation BERT model of the target from the converter, and to parse the first clause object set based on the dependency syntax tree to obtain the obligation-condition relationship pair.

[0125] Each module in the aforementioned contract clause conflict detection device may be implemented in whole or in part through software, hardware, or a combination thereof. Each module may be embedded in or independent of a processor within a computer device in the form of hardware, or may be stored in a computer device memory in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0126] In an exemplary embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 7 As shown. The computer device includes a processor, memory, an input / output interface, a communication interface, a display unit, and an input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals via wired or wireless means, and the wireless means can be achieved via Wi-Fi, mobile cellular networks, NFC (near-field communication), or other technologies. When executed by the processor, the computer program implements a method for detecting conflicts in contract terms. The display unit of the computer device is used to produce visual images and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device casing, or an external keyboard, touchpad or mouse.

[0127] Those skilled in the art will understand that Figure 7 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0128] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps of any of the above-mentioned method embodiments are implemented. The technical principles and technical effects are similar and will not be repeated here.

[0129] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of any of the following method embodiments are implemented. The technical principles and technical effects are similar and will not be repeated here.

[0130] In one embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the steps of any of the above method embodiments are implemented. The technical principles and technical effects are similar and will not be repeated here.

[0131] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0132] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiments. In particular, any reference to memory, database, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processors involved in the various embodiments provided herein may be, but are not limited to, general-purpose processors, central processing units (CPUs), graphics processing units (GPUs), digital signal processors (DSPs), programmable logic devices (PLDs), data processing logic devices based on quantum computing, and the like.

[0133] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0134] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A method for detecting conflicts in contract terms, characterized in that: The method comprises: Identify key entities in the contract to be reviewed; key entities include at least the contract subject, amount, and time point; Segment the contract text to be reviewed according to its chapter and clause numbers to obtain a first clause object set of the contract text to be reviewed; Obtaining semantic vectors and obligation-condition relationship pairs corresponding to the first clause object set; According to the semantic vectors, obligation-condition relationship pairs and key entities corresponding to the first clause object set, a conflict detection model is used to perform conflict detection on the clauses of the contract text to be reviewed to obtain a conflict detection result.

2. The method according to claim 1, characterized in that The method further comprises: Obtain historical contract texts, legal provisions, and precedents on contract disputes; Segment the historical contract text according to its chapter and clause numbers to obtain a second clause object set of the historical contract text; The second clause object set, the legal provisions, and the contract dispute precedent are respectively taken as nodes, and the relationship between the nodes is obtained; Constructing an initial knowledge graph based on the nodes and the relationships between the nodes; The conflict detection model is obtained by training an initial conflict detection model based on the initial knowledge graph.

3. The method according to claim 2, characterized in that The method further comprises: Obtaining feedback information regarding the conflict detection result; Update the initial knowledge graph according to the feedback information to obtain a new knowledge graph; The conflict detection model is trained based on the new knowledge graph to optimize the conflict detection model.

4. The method according to claim 2 or 3, characterized in that The obtaining of the relationship between the nodes includes: The conflict type and the association strength weight between the nodes are obtained, and the relationship between the nodes is determined according to the conflict type and the association strength weight between the nodes.

5. The method according to claim 1, wherein The conflict detection is performed on the clauses of the contract text to be reviewed using a conflict detection model based on the semantic vectors, obligation-condition relationship pairs, and key entities corresponding to the first clause object set, to obtain a conflict detection result, including: Determining edge features between the clause object sets according to the semantic vectors, obligation-condition relationship pairs, and key entities corresponding to the clause object set groups; the clause object set groups include any two of the first clause object sets; For each of the clause object set groups, determining the conflict probability between two first clause object sets in the clause object set group using the conflict detection model according to the edge features and corresponding semantic vectors between the clause object set groups; Based on the conflict probabilities corresponding to the clause object set groups, a conflict detection result of the contract text to be reviewed is obtained.

6. The method according to claim 1, characterized in that The obtaining of the semantic vector and obligation condition relationship pair corresponding to the first clause object set includes: The target is derived from the bidirectional encoder representation BERT model of the converter to obtain the semantic vector corresponding to the first clause object set, and the first clause object set is parsed based on the dependency syntax tree to obtain the obligation condition relationship pair.

7. A contract clause conflict detection device, characterized in that: The device comprises: An identification module is used to identify key entities in the contract text to be reviewed; the key entities include at least the contract subject, amount, and time point; A first segmentation module is configured to segment the contract text to be reviewed according to its chapters and clause numbers to obtain a first clause object set of the contract text to be reviewed; A first acquisition module is used to obtain the semantic vector and obligation condition relationship pair corresponding to the first clause object set; The detection module is used to perform conflict detection on the terms of the contract text to be reviewed using a conflict detection model based on the semantic vectors, obligation-condition relationship pairs and key entities corresponding to the first term object set to obtain conflict detection results.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images