Contract information processing method and electronic equipment

Abstract

The invention provides a contract information processing method and electronic equipment, and relates to the technical field of data processing. The method comprises the following steps: acquiring a to-be-processed contract document, and preprocessing the to-be-processed contract document to obtain a processed contract document; content extraction is conducted on the processed contract document, key performance information in the processed contract document is obtained, and the key performance information comprises multiple performance execution tasks and performance element information of the performance execution tasks; constructing a task dependency relationship graph according to the performance key information; according to the task dependency graph, generating a performance execution plan of the to-be-processed contract document; and performing visual display on the performance execution plan according to each item in the performance execution plan. According to the method and the device, the contract document is quickly and accurately analyzed, and the contract fulfillment task is effectively planned and monitored, so that the efficiency and the accuracy of executing the contract fulfillment task are greatly improved, and the requirement of continuously increasing business volume of an enterprise is met.

Description

Technical Field

[0001] This application relates to the technical field of data processing, and more particularly, to a method for processing contract information and an electronic device. Background Art

[0002] In the current business environment, the accuracy and efficiency of contract performance are crucial to the success of enterprises. However, many traditional enterprises still adopt a contract performance processing method that relies on a large number of manual operations. This processing method is not only time-consuming but also prone to information omission or misjudgment due to human errors, thus posing many challenges to the contract review process and performance execution management.

[0003] Currently, when processing various performance execution tasks in a contract, such as contract information extraction, task assignment and tracking, compliance check, and contract management and filing, the traditional method is for staff to manually extract key information from a large number of contract texts and identify and register them one by one. This process is not only inefficient and time-consuming but also increases the risk of human errors. However, with the continuous growth of enterprise business volume, this manual processing method is difficult to meet the needs of efficient operation, and the efficiency and accuracy of contract performance task execution cannot be guaranteed. Summary of the Invention

[0004] The purpose of this application is to provide a method for processing contract information and an electronic device to address the problem in the prior art that with the continuous growth of enterprise business volume, the traditional manual processing method is difficult to meet the needs of efficient operation, and the efficiency and accuracy of contract performance task execution cannot be guaranteed.

[0005] To achieve the above object, the technical solutions adopted in the embodiments of this application are as follows:

[0006] In a first aspect, an embodiment of this application provides a method for processing contract information, the method including:

[0007] Obtain a contract document to be processed, and preprocess the contract document to be processed to obtain a processed contract document, where the processed contract document includes multiple paragraphs, and each paragraph includes: paragraph serial number, soft line break position serial number, and paragraph content;

[0008] Extract the content of the processed contract document to obtain the key performance information in the processed contract document, where the key performance information includes multiple performance execution tasks and the performance element information of each performance execution task;

[0009] Construct a task dependency graph according to the key performance information, where the task dependency graph is used to indicate the execution order and dependency relationship of each performance execution task;

[0010] Generate an implementation plan for fulfilling the contract of the to-be-processed contract document according to the task dependency graph, where the implementation plan for fulfilling the contract includes: multiple matters arranged in sequence, and each matter includes: execution time and execution content;

[0011] Visualize the implementation plan for fulfilling the contract according to each matter in the implementation plan for fulfilling the contract.

[0012] As a possible implementation, the preprocessing of the to-be-processed contract document to obtain a processed contract document includes:

[0013] Perform image recognition on the to-be-processed contract document to obtain a target image, extract the text information in the target image, and use the text information in the target image to replace the target image in the to-be-processed contract document to obtain a plain text contract document;

[0014] Normalize the plain text contract document to obtain a normalized contract document, and perform structured processing on the normalized contract document to obtain the processed contract document.

[0015] As a possible implementation, the extraction of the key information for fulfilling the contract from the processed contract document includes:

[0016] Input the processed contract document into a pre-trained natural language processing model, and the natural language processing model extracts the key entities in the processed contract document according to the semantic information of the paragraph content in each paragraph of the processed contract document. The key entities at least include: fulfillment time, fulfillment method, and liability for breach of contract;

[0017] Determine each task for fulfilling the contract in the processed contract document according to the key entities and the context information corresponding to the key entities;

[0018] Perform entity extraction on the text content corresponding to each task for fulfilling the contract to obtain the fulfillment element information of each task for fulfilling the contract.

[0019] As a possible implementation, after extracting the key information for fulfilling the contract from the processed contract document, it further includes:

[0020] Classify each task for fulfilling the contract to obtain a classification result;

[0021] Based on the classification result, use a preset verification rule to verify the key information for fulfilling the contract. The preset verification rule includes: regular expression rule and multi-level verification rule.

[0022] As a possible implementation manner, based on the classification result, verifying the key performance information by using a preset verification rule includes:

[0023] Based on the classification result, determining the format type of the performance element information of the performance execution task;

[0024] Screening a regular expression rule corresponding to the format type from a preset pattern matching rule library, and verifying the format of the performance element information based on the regular expression rule.

[0025] As a possible implementation manner, constructing a task dependency graph according to the key performance information includes:

[0026] Performing dependency syntactic analysis and semantic role annotation on the text content of each performance execution task to determine multiple statement components in the text content of each performance execution task and the semantic relationships between the statement components;

[0027] Determining the dependency relationships between the performance execution tasks according to the multiple statement components in each performance execution task and the semantic relationships between the statement components;

[0028] Determining the execution order of each performance execution task according to the performance time information of each performance execution task;

[0029] Constructing the task dependency graph according to the execution order of each performance execution task and the dependency relationships between the performance execution tasks. The task dependency graph includes multiple task nodes and multiple directed edges. Each task node represents a performance execution task, and the directed edge between two task nodes indicates that there is a dependency relationship between the performance execution tasks represented by the two task nodes.

[0030] As a possible implementation manner, performing dependency syntactic analysis and semantic role annotation on the text content of each performance execution task to determine multiple statement components in the text content of each performance execution task and the semantic relationships between the statement components includes:

[0031] For each performance execution task, performing word segmentation on the text content of the performance execution task to obtain a statement sequence corresponding to the performance execution task;

[0032] Performing part-of-speech tagging on each word in the statement sequence corresponding to the performance execution task to obtain the part-of-speech tags of the words, and determining multiple statement components according to the part-of-speech tags of the words;

[0033] Based on the part-of-speech tags of each word, select target words with target part-of-speech tags and the target arguments of the target words from each word in the statement sequence, and perform semantic role annotation on the target arguments to obtain the semantic role tags of the target arguments, where the target arguments refer to the entities associated with the target words in the statement sequence;

[0034] Determine the semantic relationships between each statement component according to the part-of-speech tags of each word and the semantic role tags of the target arguments.

[0035] As a possible implementation, the determining of the dependency relationships between each performance execution task according to multiple statement components in each performance execution task and the semantic relationships between each statement component includes:

[0036] Determine the key semantic relationships between each performance execution task according to multiple statement components in each performance execution task and the semantic relationships between each statement component, where the key semantic relationships are used to represent the logical relationships between each performance execution task;

[0037] Determine the dependency relationships between each performance execution task according to the key semantic relationships between each performance execution task.

[0038] As a possible implementation, the generating of the performance execution plan for the to-be-processed contract document according to the task dependency graph includes:

[0039] Determine the time window information of each performance execution task according to the performance time information, execution order, and dependency relationships of each performance execution task, where the time window information includes the earliest start time and the latest end time of the performance execution task;

[0040] Perform resource allocation according to the time window information of each performance execution task and the current available resource information to obtain the resource allocation results of each performance execution task;

[0041] Generate the performance execution plan according to the time window information of each performance execution task and the resource allocation results.

[0042] As a possible implementation, the determining of the time window information of each performance execution task according to the performance time information, execution order, and dependency relationships of each performance execution task includes:

[0043] Convert the performance time information of each performance execution task into a preset time format and adjust the performance time information of each performance execution task according to the time zone difference to obtain the target performance time of each performance execution task;

[0044] Determine the time window information for each performance execution task according to the target performance time, the execution order, and the dependency relationship of each performance execution task.

[0045] As a possible implementation, after generating the performance execution plan for the to-be-processed contract document according to the task dependency graph, it further includes:

[0046] Determine reminder items according to the performance execution plan, and store the reminder items in an ordered set in a preset database. The ordered set includes: reminder items, identifiers of reminder items, and expiration times of reminder items;

[0047] Pull the expired target reminder items from the preset database at a preset agreed time, and obtain the user information associated with the target reminder items based on the identifier of the target reminder items;

[0048] Send a reminder notification to the target user corresponding to the user information, and delete the identifier of the target reminder item from the ordered set after successful sending. The reminder notification includes at least the item content and expiration time of the reminder item.

[0049] In a second aspect, an embodiment of the present application provides a contract information processing device, and the device includes:

[0050] An acquisition module, configured to acquire a to-be-processed contract document, and preprocess the to-be-processed contract document to obtain a processed contract document. The processed contract document includes multiple paragraphs, and each paragraph includes: paragraph number, soft line break position number, and paragraph content;

[0051] A content extraction module, configured to extract the key performance information in the processed contract document to obtain the key performance information in the processed contract document. The key performance information includes multiple performance execution tasks and the performance element information of each performance execution task;

[0052] A construction module, configured to construct a task dependency graph according to the key performance information. The task dependency graph is used to indicate the execution order and dependency relationship of each performance execution task;

[0053] A generation module, configured to generate a performance execution plan for the to-be-processed contract document according to the task dependency graph. The performance execution plan includes: multiple items arranged in sequence, and each item includes: execution time, execution content;

[0054] A visualization module, configured to visually display the performance execution plan according to each item in the performance execution plan.

[0055] In a third aspect, an embodiment of the present application provides an electronic device, including: a processor, a storage medium, and a bus. The storage medium stores machine-readable instructions executable by the processor. When the electronic device is running, the processor communicates with the storage medium through the bus, and the processor executes the machine-readable instructions to perform the steps of the contract information processing method according to any one of the above first aspects.

[0056] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium. A computer program is stored on the computer-readable storage medium. When the computer program is run by a processor, it performs the steps of the contract information processing method according to any one of the above first aspects.

[0057] According to the contract information processing method and the electronic device of the embodiments of the present application, a contract document to be processed is obtained, and the contract document to be processed is preprocessed to obtain a processed contract document; content extraction is performed on the processed contract document to obtain the key performance information in the processed contract document; according to the key performance information, a task dependency graph is constructed; according to the task dependency graph, a performance execution plan for the contract document to be processed is generated; according to each item in the performance execution plan, a visual display of the performance execution plan is performed. According to the embodiments of the present application, the contract document to be processed is obtained and preprocessed, the contract content is segmented into paragraphs with clear structures, and paragraph numbers and soft line break position numbers are marked. Then, content extraction is performed on the paragraphs to accurately identify the key performance information, including multiple performance execution tasks and their elements, greatly improving the accuracy and efficiency of information extraction. Furthermore, a task dependency graph is constructed based on the key performance information to intuitively display the logical connection and dependency relationship between various performance execution tasks, ensuring that the performance tasks are executed in the correct order. And a detailed performance execution plan is generated according to the task dependency graph, clearly listing the execution time and content of each item, making the task arrangement more scientific and reasonable. And a visual display of the performance execution plan is used, which not only improves the efficiency and accuracy of the execution of the contract performance tasks, enhances user convenience, but also meets the needs of the rapid development of the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required to be used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0059] Figure 1 FIG. shows a schematic flowchart of a contract information processing method provided by an embodiment of the present application;

[0060] Figure 2 Shows a schematic flowchart of a preprocessing method provided by an embodiment of the present application;

[0061] Figure 3 Shows a schematic flowchart of a content extraction method provided by an embodiment of the present application;

[0062] Figure 4 Shows a schematic flowchart of a rule verification method provided by an embodiment of the present application;

[0063] Figure 5 Shows a schematic flowchart of a method for constructing a task dependency graph provided by an embodiment of the present application;

[0064] Figure 6 Shows a schematic flowchart of a method for determining semantic relationships provided by an embodiment of the present application;

[0065] Figure 7 Shows a schematic diagram of a task dependency graph provided by an embodiment of the present application;

[0066] Figure 8 Shows a schematic flowchart of a method for generating a performance execution plan provided by an embodiment of the present application;

[0067] Figure 9 Shows a schematic flowchart of a performance reminder method provided by an embodiment of the present application;

[0068] Figure 10 Shows a data conversion timing diagram for performance extraction information provided by an embodiment of the present application;

[0069] Figure 11 Shows a schematic diagram of a reminder item processing interface provided by an embodiment of the present application;

[0070] Figure 12 Shows a schematic flowchart of a method for generating a contract performance plan provided by an embodiment of the present application;

[0071] Figure 13 Shows a schematic diagram of the structure of a contract information processing device provided by an embodiment of the present application;

[0072] Figure 14 Shows a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. Detailed implementation manners

[0073] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of this application. It should be understood that the accompanying drawings in this application are only for the purposes of illustration and description, and are not used to limit the protection scope of this application. Additionally, it should be understood that the schematic drawings are not drawn to actual scale. The flowcharts used in this application illustrate the operations implemented according to some embodiments of this application. It should be understood that the operations in the flowchart may not be implemented in sequence, and steps without a logical context relationship may be reversed in order or implemented simultaneously. Moreover, those skilled in the art can add one or more other operations to the flowchart or remove one or more operations from the flowchart under the guidance of the content of this application.

[0074] In addition, the described embodiments are only some embodiments of this application, rather than all embodiments. The components of the embodiments of this application described and illustrated herein generally can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative efforts fall within the protection scope of this application.

[0075] It should be noted that the term "including" will be used in the embodiments of this application to indicate the existence of the subsequently stated features, but does not exclude the addition of other features.

[0076] To improve the management efficiency of contract performance, this application uses natural language processing technology to accurately extract the contract performance plan, intelligently generate the performance execution plan, and perform visual display. This can not only significantly reduce manual intervention, improve work efficiency, but also minimize the error rate to ensure that the commitments in each contract are properly fulfilled.

[0077] Figure 1 The flowchart shows a method for processing contract information provided by an embodiment of this application. Referring to Figure 1 as shown, the method specifically includes the following steps:

[0078] S101. Obtain the contract document to be processed, and preprocess the contract document to be processed to obtain the processed contract document.

[0079] Optionally, the contract document to be processed can be a document in an easy-to-process format such as a Word document or a PDF document. If the data format of the contract document to be processed is not an easy-to-process format such as Word or PDF, the format of the contract document to be processed can be converted. For example, a format conversion tool can be used to achieve the conversion between different document formats so as to preprocess the contract document to be processed.

[0080] Optionally, the processed contract document includes multiple paragraphs, and each paragraph includes: paragraph serial number, soft line break position serial number, and paragraph content. The preprocessing of the contract document to be processed includes but is not limited to text extraction processing, data cleaning processing, paragraph division, and information annotation, etc. Among them, text extraction processing refers to extracting text from some non-pure text format content contained in the contract document to be processed, such as images, etc. Data cleaning processing refers to removing irrelevant information in the contract document to be processed, such as removing irrelevant characters such as headers, footers, and annotations. Paragraph division refers to splitting the entire contract document to be processed into multiple paragraphs, and each paragraph can be recognized according to its natural structure or defined based on specific markers such as soft line break characters. Information annotation refers to adding additional information to each paragraph, such as paragraph serial number, soft line break position serial number, etc.

[0081] S102. Extract the key performance information in the processed contract document to obtain the key performance information in the processed contract document.

[0082] Optionally, the key performance information refers to the specific matters or tasks that are stipulated in the processed contract document and that both parties of the contract must abide by and execute, as well as the important details involved in the execution process of these matters or tasks. Specifically, the key performance information includes multiple performance execution tasks and the performance element information of each performance execution task.

[0083] Optionally, the present application uses a named entity model (Named Entity Recognition, NER) to perform content extraction processing on the processed contract document to identify specific types of entities in the processed contract document, such as entities such as "date", "amount", "responsible party name", etc., and based on the identified entities and the association relationships between the entities, determine multiple performance execution tasks in the processed contract document and the performance element information included in each performance execution task.

[0084] S103. Construct a task dependency graph according to the key performance information.

[0085] Optionally, a task dependency graph is used to indicate the execution order and dependencies of each performance execution task. Constructing a task dependency graph based on the key performance information mainly visualizes the performance execution tasks extracted from the processed contract documents and their dependencies on each other, so as to clearly show the execution order of each performance execution task and the dependencies between the performance execution tasks.

[0086] Optionally, determine the performance execution tasks that need to be executed from the key performance information, and analyze whether each performance execution task depends on the completion of other performance execution tasks. For example, "buyer's payment" may depend on the completion of "buyer's inspection", and determine a reasonable execution order based on the dependencies between the performance execution tasks. For example, if performance execution task B depends on performance execution task A, then performance execution task A must be completed before performance execution task B.

[0087] Optionally, in the task dependency graph, each performance execution task is displayed as a task node, and a directed edge, such as an arrow, is used to connect two performance execution tasks with an associated relationship, indicating that the latter can only start after the former is completed. For example, there is a performance execution task A of "supplier delivers goods" and a performance execution task B of "buyer inspects goods". Since the performance execution task B of "buyer inspects goods" can only start after the performance execution task A of "supplier delivers goods" is completed, a directed edge, such as an arrow, is used to connect the task node of performance execution task A and the task node of performance execution task B in the task dependency graph, and the direction of the arrow is from the performance execution task A of "supplier delivers goods" to the performance execution task B of "buyer inspects goods". In addition, more relevant information, such as the estimated completion time, responsible party, etc., can also be marked in the task dependency graph to increase the information content and practicality of the task dependency graph.

[0088] S104. Generate a performance execution plan for the contract document to be processed according to the task dependency graph.

[0089] Optionally, the performance execution plan is the result of a detailed plan for each task in the contract. The performance execution plan includes: multiple items arranged in sequence, and each item includes: execution time and execution content. Among them, the multiple items arranged in sequence are the task sequence determined based on the task dependency graph. Each item represents a specific performance action. The execution time specifies an estimated start time and end time for each item to ensure that all tasks are completed in the correct order and within an appropriate time frame. The execution content refers to the specific work content or operation describing each item.

[0090] Optionally, based on the task dependency graph, the sequence and dependencies between the performance execution tasks can be clarified. On this basis, according to the contract requirements and actual operation conditions, execution times are allocated to each performance execution task, including setting reasonable start times and end times. Then, the specific execution details of each performance execution task are elaborated, including the responsible party, quality standards, etc., and the success completion criteria for each performance execution task are clarified, thereby generating a detailed performance execution plan. Usually, the performance execution plan lists all the performance execution tasks in chronological order, along with the corresponding execution times and execution contents of each performance execution task.

[0091] S105. Visualize the performance execution plan according to each item in the performance execution plan.

[0092] Optionally, visualizing the performance execution plan means presenting each performance execution task in the performance execution plan in an intuitive and easy-to-understand form to help users better understand the timeline of the entire performance process, the sequence of tasks, and their interrelationships, thus facilitating better planning and management.

[0093] Optionally, the visualization methods include but are not limited to Gantt charts, flowcharts, timelines, network diagrams, etc. Exemplarily, a Gantt chart is a bar chart that shows the start and end times of each item in the performance execution plan. Its horizontal axis represents time, and the vertical axis represents all items. Each item is represented by a horizontal bar, and its length corresponds to the duration of the item. A flowchart shows a series of items and the logical sequence between them through graphical symbols, clearly demonstrating the dependencies and flow directions between items. A timeline is a method of marking the time points when items occur along a straight line, intuitively showing the occurrence order and relative time positions of items. A network diagram shows the relationships between items through nodes and arrows, clearly presenting the complex dependencies and parallelism between items.

[0094] Based on this, according to the contract information processing method provided by the embodiments of the present application, the contract document to be processed is obtained and preprocessed. The contract content is segmented into paragraphs with clear structures, and paragraph numbers and soft line break position numbers are marked. Then, content extraction is performed on the paragraphs to accurately identify the key performance information, including multiple performance execution tasks and their elements, greatly improving the accuracy and efficiency of information extraction. Furthermore, a task dependency graph is constructed based on the key performance information to visually display the logical connections and dependencies between various performance execution tasks, ensuring that the performance tasks are executed in the correct order. And a detailed performance execution plan is generated according to the task dependency graph, clearly listing the execution time and content of each item, making the task arrangement more scientific and reasonable. And the visual display of the performance execution plan not only improves the efficiency and accuracy of the contract performance task execution, enhances user convenience, but also meets the needs of the enterprise's rapid development.

[0095] Figure 2 FIG. shows a schematic flowchart of a preprocessing method provided by an embodiment of the present application. Refer to Figure 2 As shown, the above step S101 preprocesses the contract document to be processed to obtain a processed contract document, which specifically includes the following steps:

[0096] S201. Perform image recognition on the contract document to be processed to obtain a target image, and extract the text information in the target image. Use the text information in the target image to replace the target image in the contract document to be processed to obtain a pure text contract document.

[0097] Optionally, if the contract document to be processed is not a pure text document and contains some image information, it is necessary to perform image recognition on the contract document to be processed to extract the text information in the image, and insert the text information of the image back into the corresponding position in the contract document to be processed according to the position and order of the image in the original contract document to be processed, so as to obtain a pure text contract document without any image elements for subsequent content extraction processing.

[0098] Optionally, the present application uses optical character recognition technology (OCR) for image recognition to extract the text information in the recognized target image. Specifically, in order to improve the accuracy of the optical character recognition technology OCR, some preprocessing can be performed on the target image first. The preprocessing includes, but is not limited to, operations such as adjusting the resolution, grayscale conversion, binarization, denoising, skew correction, and cropping. Then, the selected OCR tool is used to perform recognition operations on the preprocessed target image to extract the text information in the target image.

[0099] S202. Normalize the pure text contract document to obtain a normalized contract document, and then perform structuring on the normalized contract document to obtain a processed contract document.

[0100] Optionally, normalizing and structuring the pure text contract document is to ensure that the content of the pure text contract document is clear, easy to understand, and convenient for subsequent analysis. Exemplarily, normalization refers to converting the original pure text contract document into a standard form with a unified format and consistent content, including cleaning the text, segmenting and splitting sentences, and formatting the text. Among them, cleaning the text includes removing irrelevant characters and correcting errors. Removing irrelevant characters means removing unnecessary characters such as extra spaces, line breaks, tab characters, etc., and any non-printable characters. Correcting errors means correcting spelling mistakes, grammar errors, or recognition errors that may occur during the OCR process. Segmenting the text means dividing the text into multiple paragraphs according to logical meaning or theme, and each paragraph focuses on a central idea. Splitting sentences means accurately dividing the sentence boundaries to ensure that each sentence is a complete and independent unit of meaning. Formatting the text includes unifying the font size, style, and color, standardizing the use of punctuation marks, such as ensuring the consistency of all full stops, commas, quotation marks, etc., and adjusting typesetting details such as indentation and line spacing.

[0101] Optionally, structuring refers to further organizing the information on the basis of normalization to make it a structured data form for easy computer processing and analysis. Structuring includes operations such as information extraction, establishing a hierarchical structure, adding metadata, and annotating association relationships. Specifically, information extraction is to extract key information from the text, such as important elements like dates, amounts, party names, clause numbers, etc. Establishing a hierarchical structure means constructing a hierarchical structure according to the internal logic of the contract. For example, a contract can be divided into multiple parts such as a preamble, definition clauses, main body clauses, liability for breach of contract, and dispute resolution. Adding metadata means adding descriptive tags or metadata to the document and its various parts. For example, adding an index to each paragraph obtained by segmentation, and the index includes paragraph numbers and soft line break position numbers, etc. Annotating association relationships means annotating the relationships between the contents of various parts of the document, such as which clauses reference each other and which conditions depend on other conditions.

[0102] Based on this, by converting the contract document to be processed into a pure text contract document and performing normalization and structuring on the pure text contract document, the finally obtained processed contract document not only retains all the information of the original text but also is organized into a machine-readable form, greatly improving the information retrieval efficiency and automated processing ability.

[0103] Figure 3 shows a schematic flowchart of a content extraction method provided by an embodiment of the present application. Refer to Figure 3As shown, the above step S102 extracts the content of the processed contract document to obtain the key performance information in the processed contract document, which specifically includes the following steps:

[0104] S301. Input the processed contract document into a pre-trained natural language processing model. The natural language processing model extracts the key entities in the processed contract document according to the semantic information of the paragraph content in each paragraph of the processed contract document.

[0105] Optionally, the natural language processing model NLP is often constructed based on deep learning methods, such as pre-trained language models like BERT and RoBERTa. However, although pre-trained language models already have a certain language understanding ability, in order to better adapt to specific tasks, such as extracting key entities from contracts, it is usually necessary to fine-tune the pre-trained language model with annotated contract data to improve the understanding accuracy of the pre-trained language model for the terms and structures in the specific field of contracts.

[0106] Optionally, after inputting the processed contract document into the pre-trained natural language processing model, the natural language processing model will understand its meaning according to the content of each paragraph and its context. Specifically, it can rely on the complex neural network structure inside the natural language processing model to capture the semantic associations between words and the logical connections between sentences. Based on this understanding of the processed contract document, the natural language processing model will locate and classify the key entities in the processed contract document.

[0107] Optionally, the key entities at least include: performance time, performance method, and liability for breach of contract. Among them, the performance time refers to the time point or time period when each task should be completed, the performance method refers to the specific methods or steps for describing how to perform the contract terms, and the liability for breach of contract refers to the responsibilities and consequences stipulated when a party fails to perform its contract obligations.

[0108] S302. Determine each performance execution task in the processed contract document according to the key entities and the context information corresponding to the key entities.

[0109] Optionally, for each identified key entity, parse its context information to determine the specific performance execution task to which the key entity belongs. For example, if "before June 2025" is identified as the performance time, it is necessary to check the sentences or paragraphs near this text to determine which performance execution task this is the time requirement for, such as "The supplier shall deliver the goods before June 2025". On this basis, define each performance execution task to obtain the corresponding text content for each performance execution task. The text content usually includes the task name and related entities. Among them, the task name is, for example, "deliver goods", "make payment", and the related entities refer to all key entities related to the task, such as performance time, performance method, etc.

[0110] S303. Perform entity extraction on the text content corresponding to each performance execution task to obtain the performance element information of each performance execution task.

[0111] Optionally, for each performance execution task, NLP technology can be used to extract in detail all relevant performance element information. The performance element information includes, but is not limited to, performance time, performance method, responsible party, prerequisite conditions, liability for breach of contract, etc. Among them, the performance time refers to when the task should be completed, the performance method refers to how to complete this task, the responsible party refers to the party responsible for executing the task, the prerequisite conditions refer to the conditions required to complete this task or other pre-tasks, and the liability for breach of contract refers to the consequences that will be faced if the task fails to be completed on time and with quality.

[0112] Exemplarily, if there is a performance execution task, the corresponding text content is "Party A shall pay the balance of RMB 500,000 within 30 days after Party B completes the project acceptance. If Party A fails to make the payment on time, it shall pay liquidated damages at a rate of three ten-thousandths of the unpaid amount per day". Performing entity extraction on the text content corresponding to this performance execution task can obtain the performance execution task of "pay the balance", and the related key entities include the performance time of "within 30 days after Party B completes the project acceptance", the amount of "RMB 500,000", and the extracted performance element information includes performance time: within 30 days after Party B completes the project acceptance, responsible party: Party A, prerequisite conditions: Party B completes the project acceptance, liability for breach of contract: if Party A fails to make the payment on time, it shall pay liquidated damages at a rate of three ten-thousandths of the unpaid amount per day.

[0113] Based on this, the present application uses natural language processing technology to extract performance execution tasks and their related performance element information from contract documents, which not only improves the efficiency of contract management, but also reduces the possibility of human errors and enhances the transparency and controllability of contract execution.

[0114] Figure 4 Shows a schematic flow chart of a rule verification method provided by an embodiment of the present application. As a possible implementation, refer toFigure 4 As shown, after obtaining the key performance information in the processed contract document, the method further includes:

[0115] S401. Classify each performance execution task to obtain a classification result.

[0116] Optionally, by analyzing the specific content of each performance execution task through NLP technology, such as "deliver goods" and "make payment", the task type to which the performance execution task belongs can be determined. Further, relevant performance element information is assigned to each task type. For example, for "make payment", its performance elements may include amount, payment time, responsible party, etc. In this way, all performance execution tasks are classified according to their types to form a structured classification result.

[0117] S402. Based on the classification result, use preset verification rules to verify the key performance information.

[0118] Optionally, the preset verification rules include: regular expression rules and multi-level verification rules. In the embodiments of the present application, using the preset verification rules to verify the classified key performance information is to ensure the accuracy and consistency of the key performance information.

[0119] Optionally, based on the classification result, determine the format type of the performance element information of the performance execution task, screen out the regular expression rules corresponding to the format type from the preset pattern matching rule library, and verify the format of the performance element information based on the regular expression rules.

[0120] Exemplarily, for regular expression rule verification, based on the classification result, identify the format type of the performance element information of each performance execution task. For example, the amount field in the "make payment" task should belong to the currency format, and the time field of "deliver product" should belong to the date format. Then, screen out the regular expression rules corresponding to these format types from the preset pattern matching rule library. For example, for the date format (year-month-day), (\d{4})-(\d{2})-(\d{2}) can be used; for the amount with currency symbol, [￥€￡]?\s*\d+(,\d{3})*(.\d{2})? can be used; for the percentage format, \d+(\.\d+)?% can be used. On this basis, apply the corresponding regular expression rules to each performance element information for format verification. If the information does not conform to the expected format, it is marked as an error or a warning.

[0121] Optionally, for multi-level verification rule checking, format verification, character set verification, and syntax and semantic consistency checking are performed on the performance element information of each performance execution task. Among them, format verification is to ensure that all performance element information follows specific format specifications, such as dates, times, currencies, etc., which can be verified by applying the above regular expression rules. Character set verification refers to checking whether there are illegal or unexpected characters in the performance element information. For example, some financial reports may not allow non-ASCII characters or specific punctuation marks to appear. Syntax and semantic consistency checking means that in addition to the formal correctness, it is also necessary to ensure the logical consistency of the performance element information. For example, the payment date should not be earlier than the current date, and the amount cannot be negative, etc.

[0122] Based on this, using the preset verification rules to verify the key performance information can effectively improve the quality and reliability of the information extracted from the contract documents, and thus ensure the accuracy of subsequent data processing and decision support.

[0123] Figure 5 The flowchart of a method for constructing a task dependency graph provided by an embodiment of the present application is shown. Refer to Figure 5 As shown, the above step S103 constructs a task dependency graph according to the key performance information, which specifically includes the following steps:

[0124] S501. Perform dependency syntactic analysis and semantic role labeling on the text content of each performance execution task to determine multiple sentence components in the text content of each performance execution task and the semantic relationships between the sentence components.

[0125] Optionally, dependency syntactic analysis refers to parsing the sentence structure to determine the dependency relationships between words. Semantic role labeling (SLR) refers to labeling the arguments (i.e., the roles participating in the action) of each verb in the sentence and clarifying their positions and functions in the sentence. For example, for each verb, its agent, patient, instrument, time, and location roles can be determined.

[0126] Optionally, in the embodiment of the present application, dependency syntactic analysis and semantic role labeling technologies are adopted. Specifically, the main components of the sentence, such as the subject, predicate, and object, are found through dependency syntactic analysis, and the roles of these components in specific actions are understood through semantic role labeling. Based on this, the logical associations between contract clauses can be automatically parsed, such as the correspondence relationships between payment conditions and delivery schedules, payment and receipt types and payment and receipt methods, and payment and receipt amounts and payment and receipt ratios.

[0127] Figure 6The flowchart of a semantic relationship determination method provided by an embodiment of the present application is shown. Refer to Figure 6 As shown, the above step S501 specifically includes the following steps:

[0128] S601. For each performance execution task, perform word segmentation on the text content of the performance execution task to obtain a sequence of sentences corresponding to the performance execution task.

[0129] Exemplarily, word segmentation means decomposing the text content of the performance execution task into individual words or lexical units (tokens), and the sequence of sentences refers to the sequence of words formed after word segmentation. For example, performing word segmentation on the text content of "The supplier shall deliver product A to the buyer before April 1, 2025" can obtain the following sequence of sentences: ["supplier", "shall", "before", "April 1, 2025", "deliver", "to", "buyer", "product A"].

[0130] S602. Perform part-of-speech tagging on each word in the sequence of sentences corresponding to the performance execution task to obtain the part-of-speech tags of each word, and determine multiple sentence components according to the part-of-speech tags of each word.

[0131] Exemplarily, part-of-speech tagging means assigning a part-of-speech tag to each word (token), such as noun, verb, adjective, etc. Based on the part-of-speech tags, the main components of the sentence are identified, such as subject, predicate, object, etc. For example, continuing with the above sequence of sentences ["supplier", "shall", "before", "April 1, 2025", "deliver", "to", "buyer", "product A"], among them, "supplier" is a noun and can be used as the subject, "deliver" is a verb and can be used as the predicate, and "product A" is a noun phrase and can be used as the object.

[0132] S603. Based on the part-of-speech tags of each word, screen out the target words with target part-of-speech tags and the target arguments of the target words from the words in the sequence of sentences, and perform semantic role tagging on the target arguments to obtain the semantic role tags of the target arguments.

[0133] Exemplarily, the target part-of-speech tag is, for example, a verb. According to the part-of-speech tags of each word, the target words with the part-of-speech tag of verb are screened out from all the words. Continuing with the above sequence of sentences ["supplier", "shall", "before", "April 1, 2025", "deliver", "to", "buyer", "product A"], "deliver" is a verb and its part-of-speech tag is verb, so "deliver" can be used as a target word.

[0134] Exemplarily, on the basis of determining the target word, further determine the target argument of the target word, where the target argument refers to the entity associated with the target word in the sentence sequence. For the target word "delivery", its target arguments include "supplier", "Product A", "buyer", etc. Further, perform semantic role annotation on the target arguments to mark the specific roles of each target argument in the action. For example, the "supplier" is the agent, that is, the person who performs the action, the "Product A" is the patient, that is, the object of the action, and the "buyer" is the beneficiary. In this way, the semantic role labels of each target argument can be obtained.

[0135] S604. Determine the semantic relationships between each sentence component according to the part-of-speech tags of each word and the semantic role labels of the target arguments.

[0136] Exemplarily, based on the part-of-speech tags of each word and the semantic role labels of the target arguments, a structure reflecting the semantic relationships between each component can be constructed. For example, a semantic graph can be constructed to visually display the semantic relationships between each component. In this semantic graph, the direction of the arrow can represent the direction of the action, and additional information such as time can also be added and connected to the action in the form of independent nodes to represent the specific time limit of the action.

[0137] Based on this, by parsing the text content of the performance execution task, accurately understand the specific meaning and its components of each performance execution task, which not only improves the accuracy of data processing, but also enhances the system's ability to understand complex texts.

[0138] S502. Determine the dependency relationships between each performance execution task according to multiple sentence components in each performance execution task and the semantic relationships between each sentence component.

[0139] Optionally, determine the key semantic relationships between each performance execution task according to multiple sentence components in each performance execution task and the semantic relationships between each sentence component, and determine the dependency relationships between each performance execution task according to the key semantic relationships between each performance execution task.

[0140] Exemplarily, the key semantic relationships are used to represent the logical relationships between each performance execution task, that is, how a performance execution task specifically affects another performance execution task. Specifically, the key semantic relationships include but are not limited to: sequential relationship, parallel relationship, conditional relationship, causal relationship. Among them, the sequential relationship means that some tasks must start after other tasks are completed, the parallel relationship means that multiple tasks can be carried out simultaneously without direct dependence on each other, the conditional relationship means that the execution of a task depends on meeting specific conditions or completing specific actions, and the causal relationship means that the result of one task directly affects the start or completion of another task.

[0141] Exemplarily, analyze the logical connections between different performance execution tasks. For example, there are two performance execution tasks: "The supplier shall deliver Product A to the buyer before April 1, 2025" and "The buyer shall make payment within 30 days after receiving the goods". Based on multiple statement components in these two performance execution tasks and the semantic relationships between the statement components, it can be determined that "The buyer shall make payment within 30 days after receiving the goods" indicates that the payment behavior depends on the completion of "the supplier delivering Product A". Furthermore, it can be obtained that its logical relationships include sequential relationship and parallel relationship. The sequential relationship specifically means that the task of making payment can only be carried out after receiving the product. The conditional relationship specifically means that the prerequisite for making payment is to confirm that the product has been received and passed the inspection. Thus, the dependency relationship between the above two performance execution tasks can be obtained. Similarly, the dependency relationships between each performance execution task can be obtained respectively.

[0142] S503. Determine the execution order of each performance execution task according to the performance time information of each performance execution task.

[0143] Exemplarily, extract clear performance time information from each performance execution task. The performance time information includes but is not limited to specific dates, time periods, or times relative to other events. For example, in "The supplier shall deliver Product A before April 1, 2025", it contains a clear time point, that is, the performance time information is direct time. In the case of "The buyer shall make payment within 30 days after receiving the goods", it is necessary to first identify the trigger condition, that is, "receiving the goods", and then calculate the corresponding subsequent time, that is, the performance time information is relative time.

[0144] Exemplarily, in order to facilitate the comparison of performance times between different tasks, it is necessary to convert all time information into a unified standard format. For example, convert all time points into the standard date-time format, or convert relative time into a specific date relative to a certain reference time. On this basis, sort all performance execution tasks by combining time information and dependency relationships. Specifically, first arrange those tasks without mutual dependencies in chronological order, and then process tasks with dependency relationships to ensure that each task is arranged only after its preceding task is completed. Specifically, a topological sorting algorithm can be used to process the execution order between performance execution tasks with dependency relationships.

[0145] S504. Construct a task dependency graph according to the execution order of each performance execution task and the dependency relationships between each performance execution task.

[0146] Optionally, the task dependency graph includes multiple task nodes and multiple directed edges. Each task node represents a performance execution task respectively, and the directed edge between two task nodes indicates that there is a dependency relationship between the performance execution tasks represented by the two task nodes.

[0147] Exemplarily, referring to Figure 7 the shown task dependency graph, which includes five performance execution tasks T1, T2, T3, T4, and T5. There are two directed edges starting from T1, one pointing to T2 and the other pointing to T3, indicating that both T2 and T3 need to wait for T1 to complete. There is a directed edge from T2 to T4, and at the same time, there is also a directed edge from T3 to T4, showing that T4 can start only after both T2 and T3 are completed. Additionally, there is a separate directed edge from T3 to T5, indicating that T5 can start as long as T3 is completed. It can be seen that in Figure 7 the shown task dependency graph, the dependency relationships among these five performance execution tasks are as follows: T1: no pre-task; T2: can start only after T1 is completed; T3: can start only after T1 is completed; T4: can start only after both T2 and T3 are completed; T5: can start only after T3 is completed.

[0148] Based on this, by creating nodes (representing tasks) and connecting these nodes with arrows (directed edges) to show the dependency relationships between tasks, a task dependency graph is generated to clearly display the task execution order and dependency relationships in the entire contract document.

[0149] Figure 8 The flowchart shows a method for generating a performance execution plan provided by an embodiment of the present application. Referring to Figure 8 as shown, the above step S104 generates a performance execution plan for the contract document to be processed according to the task dependency graph, which specifically includes the following steps:

[0150] S801. Determine the time window information for each performance execution task according to the performance time information, execution order, and dependency relationships of each performance execution task.

[0151] Optionally, the time window information includes the earliest start time and the latest end time for the performance execution task. Among them, the earliest start time for performance is usually the earliest date when the task can start, which may be based on the contract effective date or other preconditions. The latest end time for performance refers to the latest date when the task must be completed, which is usually clearly specified in the contract or derived according to the dependency relationships.

[0152] Specifically, format conversion is performed on the performance time information of each performance execution task according to the preset time format, and the performance time information of each performance execution task is adjusted according to the time zone difference to obtain the target performance time for each performance execution task; according to the target performance time, execution order, and dependency relationships of each performance execution task, the time window information for each performance execution task is determined.

[0153] Exemplarily, since the performance time information of each performance execution task may be represented in different formats, it is necessary to convert various date representations into precise calendar dates. For example, for a direct date such as "April 1, 2025", it can be directly converted, while for a relative date such as "the first working day of next month", the current month can be obtained and the first day of the next month can be calculated, and then the first non-weekend working day can be found. For a relative date such as "within 60 days after the contract comes into effect", the specific due date needs to be calculated based on the contract effective date plus 60 days.

[0154] Exemplarily, for periodic performance execution tasks such as regular payments or other repetitive activities, according to the periodicity of the performance execution tasks, such as daily, weekly, monthly, quarterly, etc., the specific start and end dates of each period are automatically calculated. For time zone and holiday adjustments, the time zone differences and legal holiday arrangements in different regions can be considered to ensure the global applicability and compliance of the calculation results. Among them, regarding time zone differences, the time zones where each party is located can be determined, and the date and time can be adjusted as needed to adapt to different time zones. Regarding holiday adjustments, considering the impact of holidays on the business, it is ensured that the calculated dates do not fall on holidays. In summary, format conversion and time zone adjustment are performed on all the extracted performance time information, and all the extracted performance time information is converted into a unified standard format and application time zone, so as to accurately determine the time window information of each performance execution task.

[0155] S802. Perform resource allocation based on the time window information of each performance execution task and the current available resource information to obtain the resource allocation results of each performance execution task.

[0156] Exemplarily, in combination with the current available resource information, including human resources and physical resources, a mathematical modeling method is used to reasonably allocate resources for each performance execution task. Among them, the mathematical modeling methods include linear programming, genetic algorithms, etc. Linear programming is applicable to situations where the resource allocation problem is relatively simple and can be used to minimize or maximize a certain objective function (such as cost minimization or efficiency maximization) while satisfying a series of constraints (such as time limits, resource availability), while genetic algorithms are applicable to more complex problems, especially when there are a large number of variables and complex constraints, and find the optimal solution by simulating the process of natural selection. Regarding which specific mathematical modeling method to choose for resource allocation, it can be determined according to the actual application situation and is not specifically limited here.

[0157] S803. Generate a performance execution plan based on the time window information of each performance execution task and the resource allocation results.

[0158] Exemplarily, according to the time window information of each compliance execution task and the resource allocation result, a compliance execution plan including a plan structure of task list, schedule, personnel arrangement, and resource arrangement can be generated. Among them, the task list lists all tasks to be completed and their detailed descriptions, the schedule includes the specific start time and end time of each task, the personnel arrangement includes the execution personnel corresponding to each task, and the resource arrangement includes the specific tools and materials required for each task.

[0159] Based on this, resource allocation based on the current available resource information can not only efficiently utilize the available resources, but also ensure that each compliance execution task is completed on time and with high quality. In this way, through reasonable resource allocation and detailed compliance execution plans, the efficiency and success rate of contract compliance management can be significantly improved, and problems such as compliance delays or cost overruns caused by insufficient resources or improper scheduling can be reduced.

[0160] Figure 9 The flowchart of a compliance reminder method provided by an embodiment of the present application is shown. As a possible implementation manner, referring to Figure 9 As shown, after generating a compliance execution plan for the contract document to be processed according to the task dependency graph, the method further includes:

[0161] S901. Determine reminder items according to the compliance execution plan, and store the reminder items in an ordered set in a preset database.

[0162] Optionally, the ordered set includes: reminder items, identifiers of reminder items, and expiration times of reminder items. Extract key time nodes or events that need to be reminded from the compliance execution plan. Key nodes are important time points such as delivery dates and payment deadlines in the contract, and set the time for advance notice according to business requirements, such as sending a reminder 3 days before the task expires.

[0163] Exemplarily, referring to Figure 10 As shown, an embodiment of the present application also provides a visual compliance task display system. The system includes a front end and a back end, and can present key links, task progress, and key information points in the compliance process to users in a graphical and dynamic form with very high intuitiveness and accuracy. Specifically, after content extraction is performed based on a natural language processing model to obtain an extraction result, the back end can integrate the extraction result and display the integrated result on the front-end page for users to view and modify, and store the modified data in a preset database after the user modification is completed. Some reminder items can also be saved in the preset database to pull reminder items at preset agreed times for reminder.

[0164] S902. Pull the target reminder items that are due from the preset database according to the preset agreed time, and obtain the user information associated with the target reminder items based on the identifiers of the target reminder items.

[0165] Optionally, the preset agreed time is, for example, twelve o'clock every day, and it can be specifically set according to the actual situation. In the embodiment of the present application, a script can be run regularly to query all unprocessed reminder items before the current time. Since the representation of each reminder item is a unique identifier that can uniquely distinguish different reminder items, and the ID of each reminder item is associated with the operator who processes the reminder item, after pulling the target reminder items that are about to expire from the preset database, the user information associated with the target reminder items can be obtained according to the identifier ID of the target reminder items. This user information is, for example, the contact information of the operator who processes the target reminder item.

[0166] S903. Send a reminder notification to the target user corresponding to the user information, and delete the identifier of the target reminder item from the ordered set after the sending is successful.

[0167] Exemplarily, as shown in Figure 11 shown, Figure 11 two reminder items are schematically shown. After pulling a reminder item, the user information of the target user corresponding to the reminder item can be queried from the database, and a reminder notification is sent to the target user. The reminder notification includes at least the content of the reminder item and the due time. And after confirming that the notification is sent successfully, the corresponding reminder item identifier is removed from the ordered set.

[0168] Based on this, by regularly pulling the reminder items that are due, sending notifications to relevant users, and cleaning up the reminder items after processing, the present application not only ensures the efficiency and accuracy of the reminder work, but also helps to complete various performance tasks in a timely manner.

[0169] Figure 12 shows a schematic flowchart of a contract performance plan generation method provided by an embodiment of the present application. As shown in Figure 12 shown, for the contract document to be processed, the contract document is segmented into multiple content paragraphs, and then each paragraph is semantically segmented and content classified, and the key information for performance is identified through rule verification. And the extracted date data is processed and rule calculated to generate a performance plan, and finally the performance content is displayed in a visual manner. The entire contract performance plan generation process is realized by automated means, improving the efficiency and accuracy of contract performance management, and ensuring that key performance information can be identified and applied in a timely and accurate manner.

[0170] Based on this, in the embodiments of the present application, by extracting the contract performance plan, intelligent generation of relevant execution plan tasks can not only significantly reduce manual intervention, improve work efficiency, but also minimize the error rate to ensure that the commitments in each contract can be properly fulfilled.

[0171] Based on the same inventive concept, in the embodiments of the present application, there is also provided a contract information processing device corresponding to the contract information processing method. Since the principle of solving problems by the contract information processing device in the embodiments of the present application is similar to that of the above-mentioned contract information processing method in the embodiments of the present application, the implementation of the contract information processing device can refer to the implementation of the method, and the repeated parts will not be elaborated.

[0172] Refer to Figure 13 As shown, it is a schematic structural diagram of a contract information processing device provided by an embodiment of the present application. The contract information processing device 1300 includes: an acquisition module 1301, a content extraction module 1302, a construction module 1303, a generation module 1304, and a visualization module 1305, where:

[0173] The acquisition module 1301 is configured to acquire a contract document to be processed and preprocess the contract document to be processed to obtain a processed contract document. The processed contract document includes multiple paragraphs, and each paragraph includes: a paragraph serial number, a soft line break position serial number, and paragraph content;

[0174] The content extraction module 1302 is configured to extract the content of the processed contract document to obtain the key performance information in the processed contract document. The key performance information includes multiple performance execution tasks and the performance element information of each performance execution task;

[0175] The construction module 1303 is configured to construct a task dependency graph according to the key performance information. The task dependency graph is used to indicate the execution order and dependency relationship of each performance execution task;

[0176] The generation module 1304 is configured to generate a performance execution plan for the contract document to be processed according to the task dependency graph. The performance execution plan includes: a plurality of items arranged in sequence, and each item includes: an execution time and an execution content;

[0177] The visualization module 1305 is configured to visually display the performance execution plan according to each item in the performance execution plan.

[0178] Based on this, the contract information processing device according to the embodiments of the present application acquires and preprocesses the contract document to be processed, divides the contract content into paragraphs with clear structures, marks the paragraph numbers and the soft line break position numbers, and then extracts the content of the paragraphs to accurately identify the key performance information, including multiple performance execution tasks and their elements, greatly improving the accuracy and efficiency of information extraction. Furthermore, a task dependency graph is constructed based on the key performance information to visually display the logical connections and dependencies between the various performance execution tasks, ensuring that the performance tasks are executed in the correct order. And a detailed performance execution plan is generated according to the task dependency graph, clearly listing the execution time and content of each item, making the task arrangement more scientific and reasonable. And the visual display of the performance execution plan not only improves the efficiency and accuracy of the contract performance task execution, enhances the user convenience, but also meets the needs of the enterprise's rapid development.

[0179] In a possible implementation manner, the above-mentioned acquisition module 1301 is specifically configured to:

[0180] Perform image recognition on the contract document to be processed to obtain a target image, extract the text information in the target image, and use the text information in the target image to replace the target image in the contract document to be processed to obtain a pure text contract document;

[0181] Perform normalization processing on the pure text contract document to obtain a normalized contract document, and perform structured processing on the normalized contract document to obtain a processed contract document.

[0182] In a possible implementation manner, the above-mentioned content extraction module 1302 is specifically configured to:

[0183] Input the processed contract document into a pre-trained natural language processing model, and the natural language processing model extracts the key entities in the processed contract document according to the semantic information of the paragraph content in each paragraph of the processed contract document. The key entities at least include: performance time, performance method, liability for breach of contract;

[0184] Determine the various performance execution tasks in the processed contract document according to the key entities and the context information corresponding to the key entities;

[0185] Perform entity extraction on the text content corresponding to the various performance execution tasks to obtain the performance element information of the various performance execution tasks.

[0186] In a possible implementation manner, the above-mentioned content extraction module 1302 is further configured to:

[0187] Classify the various performance execution tasks to obtain a classification result;

[0188] Based on the classification results, verify the key performance information using preset verification rules, where the preset verification rules include: regular expression rules and multi-level verification rules.

[0189] In a possible implementation manner, the above-mentioned content extraction module 1302 is further configured to:

[0190] Based on the classification results, determine the format type of the performance elements information of the performance execution tasks;

[0191] Screen the regular expression rules corresponding to the format type from the preset pattern matching rule library, and verify the format of the performance elements information based on the regular expression rules.

[0192] In a possible implementation manner, the above-mentioned construction module 1303 is specifically configured to:

[0193] Perform dependency syntactic analysis and semantic role annotation on the text content of each performance execution task to determine multiple statement components in the text content of each performance execution task and the semantic relationships between the statement components;

[0194] According to the multiple statement components in each performance execution task and the semantic relationships between the statement components, determine the dependency relationships between the performance execution tasks;

[0195] According to the performance time information of each performance execution task, determine the execution order of each performance execution task;

[0196] Construct a task dependency graph according to the execution order of each performance execution task and the dependency relationships between the performance execution tasks. The task dependency graph includes multiple task nodes and multiple directed edges. Each task node represents a performance execution task, and the directed edge between two task nodes indicates that there is a dependency relationship between the performance execution tasks represented by the two task nodes.

[0197] In a possible implementation manner, the above-mentioned construction module 1303 is specifically configured to:

[0198] For each performance execution task, perform word segmentation on the text content of the performance execution task to obtain a statement sequence corresponding to the performance execution task;

[0199] Perform part-of-speech tagging on each word in the statement sequence corresponding to the performance execution task to obtain the part-of-speech tags of each word, and determine multiple statement components according to the part-of-speech tags of each word;

[0200] Based on the part-of-speech tags of each word, select the target words with the target part-of-speech tags and the target arguments of the target words from each word in the sentence sequence, and perform semantic role labeling on the target arguments to obtain the semantic role tags of the target arguments, where the target arguments refer to the entities associated with the target words in the sentence sequence;

[0201] Determine the semantic relationships between the sentence components according to the part-of-speech tags of each word and the semantic role tags of the target arguments.

[0202] In a possible implementation manner, the above-mentioned construction module 1303 is specifically used for:

[0203] Determine the key semantic relationships between the performance execution tasks according to multiple sentence components in each performance execution task and the semantic relationships between the sentence components, where the key semantic relationships are used to represent the logical relationships between the performance execution tasks;

[0204] Determine the dependency relationships between the performance execution tasks according to the key semantic relationships between the performance execution tasks.

[0205] In a possible implementation manner, the above-mentioned generation module 1304 is specifically used for:

[0206] Determine the time window information of each performance execution task according to the performance time information, execution order and dependency relationships of each performance execution task, where the time window information includes the earliest start time of performance execution of the performance execution task and the latest end time of performance execution;

[0207] Perform resource allocation according to the time window information of each performance execution task and the current available resource information to obtain the resource allocation results of each performance execution task;

[0208] Generate a performance execution plan according to the time window information of each performance execution task and the resource allocation results.

[0209] In a possible implementation manner, the above-mentioned generation module 1304 is specifically used for:

[0210] Convert the performance time information of each performance execution task according to the preset time format, and adjust the performance time information of each performance execution task according to the time zone difference to obtain the target performance time of each performance execution task;

[0211] Determine the time window information of each performance execution task according to the target performance time, execution order and dependency relationships of each performance execution task.

[0212] In a possible implementation manner, the above-mentioned generation module 1304 is further used for:

[0213] According to the performance execution plan, reminder items are determined and stored in an ordered set in a preset database. The ordered set includes: reminder items, identifiers of reminder items, and expiration times of reminder items;

[0214] Pull the target reminder items that have expired from the preset database according to the preset agreed time, and obtain the user information associated with the target reminder items based on the identifiers of the target reminder items;

[0215] Send a reminder notification to the target user corresponding to the user information, and delete the identifier of the target reminder item from the ordered set after successful sending. The reminder notification includes at least the content of the reminder item and the expiration time.

[0216] Descriptions of the processing flows of each module in the device and the interaction flows between modules can refer to the relevant descriptions in the above method embodiments and will not be elaborated here.

[0217] An embodiment of the present application also provides an electronic device 1400, as Figure 14 shown, is a schematic structural diagram of the electronic device 1400 provided by the embodiment of the present application, including: a processor 1401, a memory 1402. Optionally, a bus 1403 may also be included. The memory 1402 stores machine-readable instructions executable by the processor 1401. When the electronic device 1400 runs, the processor 1401 communicates with the memory 1402 through the bus 1403. When the machine-readable instructions are executed by the processor 1401, the steps of the contract information processing method described in any one of the above are executed.

[0218] An embodiment of the present application also provides a computer-readable storage medium. A computer program is stored on the computer-readable storage medium. When the computer program is run by a processor, the steps of the contract information processing method described in any one of the above are executed.

[0219] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described systems and devices can refer to the corresponding processes in the method embodiments and will not be elaborated in the present application. In several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. The device embodiments described above are only illustrative. For example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation. For another example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point, the displayed or discussed couplings or direct couplings or communication connections to each other can be through some communication interfaces. The indirect couplings or communication connections of the devices or modules can be electrical, mechanical, or other forms.

[0220] In addition, each functional unit in the various embodiments of the present application may be integrated into one processing unit, may exist physically alone for each unit, or two or more units may be integrated into one unit. If the function is implemented in the form of a software functional unit and sold or used as an independent product, it may be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, may be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0221] The above are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed in the present application can easily think of changes or substitutions, which should all be covered by the protection scope of the present application.

Claims

1. A method for processing contract information, characterized in that: include: Acquire a contract document to be processed, and pre-process the contract document to obtain a processed contract document, wherein the processed contract document includes a plurality of paragraphs, each paragraph including: a paragraph number, a soft line break position number, and paragraph content; Extracting the content of the processed contract document to obtain key performance information in the processed contract document, wherein the key performance information includes multiple performance execution tasks and performance element information of each performance execution task; Based on the key performance information, a task dependency graph is constructed, wherein the task dependency graph is used to indicate the execution order and dependency of each performance execution task; Generate a performance execution plan for the pending contract document according to the task dependency graph, wherein the performance execution plan includes: a plurality of items arranged in sequence, each item including: execution time and execution content; The performance execution plan is visualized based on each item in the performance execution plan.

2. The method according to claim 1, characterized in that The preprocessing of the contract document to be processed to obtain a processed contract document includes: Performing image recognition on the contract document to be processed to obtain a target image, extracting text information in the target image, and replacing the target image in the contract document to be processed with the text information in the target image to obtain a plain text contract document; The plain text contract document is normalized to obtain a normalized contract document, and the normalized contract document is structured to obtain the processed contract document.

3. The method according to claim 1, characterized in that The extracting content of the processed contract document to obtain key performance information in the processed contract document includes: Inputting the processed contract document into a pre-trained natural language processing model, the natural language processing model extracting key entities in the processed contract document according to semantic information of the content of each paragraph in the processed contract document, the key entities at least including: performance time, performance method, and liability for breach of contract; Determining various contract performance execution tasks in the processed contract document according to the key entity and the context information corresponding to the key entity; Entity extraction is performed on the text content corresponding to each contract performance execution task to obtain the performance element information of each contract performance execution task.

4. The method according to claim 1, characterized in that: After extracting the content of the processed contract document to obtain the key performance information in the processed contract document, the method further includes: Classify various contract performance execution tasks and obtain classification results; Based on the classification result, the key performance information is verified using preset verification rules, and the preset verification rules include: regular expression rules and multi-level verification rules.

5. The method according to claim 4, characterized in that Based on the classification result, verifying the key performance information using preset verification rules includes: Based on the classification result, determine the format type of the performance element information of the performance execution task; A regular expression rule corresponding to the format type is filtered from a preset pattern matching rule library, and the format of the performance element information is verified based on the regular expression rule.

6. The method according to claim 1, characterized in that The step of constructing a task dependency graph according to the key performance information includes: Perform dependency syntactic analysis and semantic role labeling on the text content of each contract performance execution task, and determine multiple sentence components in the text content of each contract performance execution task and the semantic relationship between the sentence components; Determine the dependency relationship between various contract performance execution tasks based on multiple sentence components in each contract performance execution task and the semantic relationship between the sentence components; Determine the execution order of each contract performance execution task based on the performance time information of each contract performance execution task; According to the execution order of each contract performance task and the dependency relationship between each contract performance task, the task dependency graph is constructed. The task dependency graph includes multiple task nodes and multiple directed edges. Each task node represents a contract performance task. The directed edge between two task nodes indicates that there is a dependency relationship between the contract performance tasks represented by the two task nodes.

7. The method according to claim 6, characterized in that The dependency syntactic analysis and semantic role labeling of the text content of each contract performance execution task are performed to determine multiple sentence components in the text content of each contract performance execution task and the semantic relationship between the sentence components, including: For each contract performance execution task, the text content of the contract performance execution task is segmented to obtain a sentence sequence corresponding to the contract performance execution task; Performing part-of-speech tagging on each word in the sentence sequence corresponding to the performance execution task to obtain a part-of-speech tag for each word, and determining a plurality of sentence components according to the part-of-speech tag for each word; Based on the part-of-speech tags of each word, a target word with a target part-of-speech tag and a target argument of the target word are selected from each word in the sentence sequence, and the target argument is annotated with a semantic role to obtain a semantic role label of the target argument, wherein the target argument refers to an entity associated with the target word in the sentence sequence; The semantic relationship between the sentence components is determined based on the part-of-speech tag of each word and the semantic role tag of the target argument.

8. The method according to claim 6, characterized in that Determining the dependency relationship between various contract performance execution tasks according to multiple sentence components in various contract performance execution tasks and the semantic relationship between the sentence components includes: Determine the key semantic relationship between the various contract performance execution tasks based on multiple sentence components in the various contract performance execution tasks and the semantic relationship between the sentence components, wherein the key semantic relationship is used to represent the logical relationship between the various contract performance execution tasks; Determine the dependency relationship between various contract fulfillment execution tasks based on the key semantic relationship between them.

9. The method according to claim 1, characterized in that: Generating a contract performance execution plan for the contract document to be processed according to the task dependency graph includes: Determine the time window information of each performance execution task according to the performance time information, execution order and dependency relationship of each performance execution task, wherein the time window information includes the earliest performance start time and the latest performance end time of the performance execution task; Allocate resources according to the time window information of each contract execution task and the currently available resource information to obtain the resource allocation results of each contract execution task; The contract execution plan is generated based on the time window information of each contract execution task and the resource allocation results.

10. The method according to claim 9, characterized in that Determining the time window information of each contract performance execution task according to the performance time information, execution order and dependency relationship of each contract performance execution task includes: The performance time information of each performance execution task is converted into a format according to a preset time format, and the performance time information of each performance execution task is adjusted according to the time zone difference to obtain the target performance time of each performance execution task; The time window information of each contract fulfillment execution task is determined according to the target fulfillment time of each contract fulfillment execution task, the execution order and the dependency relationship.

11. The method according to claim 1, characterized in that After generating the performance execution plan of the pending contract document according to the task dependency graph, the method further includes: Determine reminder items according to the contract performance execution plan, and store the reminder items in an ordered set in a preset database, wherein the ordered set includes: reminder items, reminder item identifiers, and reminder item expiration dates; Pulling out the due target reminder items from the preset database according to the preset agreed time, and acquiring the user information associated with the target reminder items based on the identifiers of the target reminder items; A reminder notification is sent to the target user corresponding to the user information, and the identifier of the target reminder item is deleted from the ordered set after the sending is successful, wherein the reminder notification at least includes the item content and the expiration time of the reminder item.

12. An electronic device, characterized in that: include: A processor and a memory, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor executes the machine-readable instructions to perform the steps of the contract information processing method as described in any one of claims 1 to 11.

Images