CN114549241A - Contract examination method, device, system and computer readable storage medium - Google Patents
Contract examination method, device, system and computer readable storage medium Download PDFInfo
- Publication number
- CN114549241A CN114549241A CN202210165386.6A CN202210165386A CN114549241A CN 114549241 A CN114549241 A CN 114549241A CN 202210165386 A CN202210165386 A CN 202210165386A CN 114549241 A CN114549241 A CN 114549241A
- Authority
- CN
- China
- Prior art keywords
- contract
- examination
- text
- model
- treaty
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/18—Legal services; Handling legal documents
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
- G06F40/295—Named entity recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
Abstract
The invention discloses a contract examination method, a device, a system and a computer readable storage medium, wherein the method comprises the following steps: receiving a contract text, inputting the contract text into a text matching model to obtain a template difference result, and inputting the contract text into an entity naming identification model to extract an element set in the contract text; and inputting the contract text into a reading understanding model to obtain an examination term analysis result, and obtaining a contract examination result according to a preset expert knowledge base, the template difference result, the element set and the examination term analysis result. According to the method, the contract text is processed through the text matching model, the entity naming recognition model and the reading understanding model, and the contract examination result is obtained by combining the preset expert knowledge base, so that the efficiency and the accuracy of the contract examination are improved.
Description
Technical Field
The invention relates to the technical field of financial science and technology, in particular to a contract examination method, a contract examination device, a contract examination system and a computer-readable storage medium.
Background
With the development of intra-row business, currently, tens of thousands of contracts need to be inspected every year, and at present, the contracts are inspected manually, but the requirements of the contract inspection on the professional ability of inspectors are high, and the flow of the contract inspection is complex, so that the efficiency and the accuracy of manual inspection of the contracts are low. Therefore, how to improve the efficiency and accuracy of contract examination is an urgent problem to be solved.
Disclosure of Invention
The invention mainly aims to provide a contract examination method, a contract examination device, a contract examination system and a computer readable storage medium, and aims to solve the problem of how to improve the efficiency and accuracy of contract examination.
To achieve the above object, the present invention provides a contract inspection method, comprising the steps of:
receiving a contract text, inputting the contract text into a text matching model to obtain a template difference result, and inputting the contract text into an entity naming identification model to extract an element set in the contract text;
and inputting the contract text into a reading understanding model to obtain an examination term analysis result, and obtaining a contract examination result according to a preset expert knowledge base, the template difference result, the element set and the examination term analysis result.
Preferably, the step of inputting the contract text into a text matching model to obtain a template difference result comprises:
inputting the contract text into a text matching model, determining the type of the contract text through the text matching model, and determining a standard contract template according to the type of the contract text;
and comparing the contract text with the standard contract template through the text matching model to obtain a template difference result.
Preferably, the step of inputting the contract text into an entity naming recognition model to extract the element set in the contract text comprises:
inputting the contract text into an entity naming recognition model, recognizing and classifying each element in the contract text through the entity naming recognition model, determining position information corresponding to each element, and extracting an element set in the contract text according to the position information.
Preferably, the step of inputting the contract text into the reading understanding model to obtain the examination term analysis result comprises:
inputting the contract text into a reading understanding model, determining the type of the contract text through the reading understanding model, and determining examination terms according to the type of the contract text;
and identifying the contract text through the reading understanding model to obtain a logic problem, and obtaining an analysis result of the examination clause according to the logic problem and the examination clause.
Preferably, the step of obtaining the contract examination result according to the preset expert knowledge base, the template difference result, the element set and the examination term analysis result comprises:
determining non-compliant elements in the element set according to a preset expert knowledge base, the template difference result, the element set and the examination term analysis result, and determining examination opinions corresponding to the non-compliant elements;
and determining the position information of the non-compliance element in the contract text, and annotating the contract text according to the position information and the examination opinions to obtain a contract examination result.
Preferably, before the step of receiving the contract text, the contract inspection method includes:
acquiring historical contract examination data, and training different sub-models in a pre-training model according to the historical contract examination data to respectively obtain a text matching pre-model, an entity naming recognition pre-model and a reading understanding pre-model;
and acquiring verification contract inspection data, inputting the verification contract inspection data into the text matching pre-model, the entity naming identification pre-model and the reading understanding pre-model respectively to obtain verification results, and determining the text matching model, the entity naming identification model and the reading understanding model based on the verification results.
Preferably, after the step of obtaining the contract examination result according to the preset expert knowledge base, the template difference result, the element set and the examination term analysis result, the contract examination method includes:
and updating historical contract examination data according to the contract examination result, and performing optimization operation on the text matching model, the entity naming recognition model and the reading understanding model according to the updated historical contract examination data.
Further, to achieve the above object, the present invention also provides a contract inspection apparatus comprising:
the extraction module is used for receiving a contract text, inputting the contract text into a text matching model to obtain a template difference result, and inputting the contract text into an entity naming identification model to extract an element set in the contract text;
and the examination module is used for inputting the contract text into a reading understanding model to obtain an examination term analysis result and obtaining a contract examination result according to a preset expert knowledge base, the template difference result, the element set and the examination term analysis result.
Preferably, the extraction module is further configured to:
inputting the contract text into a text matching model, determining the type of the contract text through the text matching model, and determining a standard contract template according to the type of the contract text;
and comparing the contract text with the standard contract template through the text matching model to obtain a template difference result.
Preferably, the extraction module is further configured to:
inputting the contract text into an entity naming recognition model, recognizing and classifying each element in the contract text through the entity naming recognition model, determining position information corresponding to each element, and extracting an element set in the contract text according to the position information.
Preferably, the extraction module is further configured to:
inputting the contract text into a reading understanding model, determining the type of the contract text through the reading understanding model, and determining an examination term according to the type of the contract text;
and identifying the contract text through the reading understanding model to obtain a logic problem, and obtaining an analysis result of the examination clause according to the logic problem and the examination clause.
Preferably, the extraction module is further configured to:
determining non-compliant elements in the element set according to a preset expert knowledge base, the template difference result, the element set and the examination term analysis result, and determining examination opinions corresponding to the non-compliant elements;
and determining the position information of the non-compliance element in the contract text, and annotating the contract text according to the position information and the examination opinions to obtain a contract examination result.
Preferably, the extraction module further comprises a training module configured to:
acquiring historical contract examination data, and training different sub-models in a pre-training model according to the historical contract examination data to respectively obtain a text matching pre-model, an entity naming recognition pre-model and a reading understanding pre-model;
and acquiring verification contract inspection data, inputting the verification contract inspection data into the text matching pre-model, the entity naming identification pre-model and the reading understanding pre-model respectively to obtain verification results, and determining the text matching model, the entity naming identification model and the reading understanding model based on the verification results.
Preferably, the review module further comprises an update module configured to:
and updating historical contract examination data according to the contract examination result, and performing optimization operation on the text matching model, the entity naming recognition model and the reading understanding model according to the updated historical contract examination data.
Further, to achieve the above object, the present invention also provides a contract inspection system including: a memory, a processor, and a treaty examination program stored on the memory and executable on the processor, the treaty examination program when executed by the processor implementing the steps of the treaty examination method as described above.
Further, to achieve the above object, the present invention also provides a computer-readable storage medium having stored thereon a contract inspection program that, when executed by a processor, implements the steps of the contract inspection method as described above.
The contract examination method provided by the invention comprises the steps of receiving a contract text, inputting the contract text into a text matching model to obtain a template difference result, and inputting the contract text into an entity naming identification model to extract an element set in the contract text; and inputting the contract text into a reading understanding model to obtain an examination term analysis result, and obtaining a contract examination result according to a preset expert knowledge base, the template difference result, the element set and the examination term analysis result. According to the method, the contract text is processed through the text matching model, the entity naming recognition model and the reading understanding model, and the contract examination result is obtained by combining the preset expert knowledge base, so that the efficiency and the accuracy of the contract examination are improved.
Drawings
FIG. 1 is a schematic diagram of an apparatus architecture of a hardware operating environment according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating a first embodiment of a contract inspection method according to the present invention.
The implementation, functional features and advantages of the objects of the present invention will be further described with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As shown in fig. 1, fig. 1 is a schematic device structure diagram of a hardware operating environment according to an embodiment of the present invention.
The device of the embodiment of the invention can be a PC or a server device.
As shown in fig. 1, the apparatus may include: a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, a communication bus 1002. The communication bus 1002 is used to implement connection communication among these components. The user interface 1003 may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface). The memory 1005 may be a high-speed RAM memory or a non-volatile memory (e.g., a magnetic disk memory). The memory 1005 may alternatively be a storage device separate from the processor 1001.
Those skilled in the art will appreciate that the configuration of the apparatus shown in fig. 1 is not intended to be limiting of the apparatus and may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components.
As shown in fig. 1, a memory 1005, which is a kind of computer storage medium, may include therein an operating system, a network communication module, a user interface module, and a contract examination program.
The operating system is a program for managing and controlling the portable contract examination system and software resources, and supports the operation of the network communication module, the user interface module, the contract examination program and other programs or software; the network communication module is used for managing and controlling the network interface 1002; the user interface module is used to manage and control the user interface 1003.
In the treaty examination system shown in fig. 1, the treaty examination system calls a treaty examination program stored in the memory 1005 by the processor 1001, and performs the operations in the various embodiments of the treaty examination method described below.
Based on the hardware structure, the embodiment of the contract examination method is provided.
Referring to fig. 2, fig. 2 is a schematic flow chart of a first embodiment of a contract examination method of the present invention, which includes:
step S10, receiving a contract text, inputting the contract text into a text matching model to obtain a template difference result, and inputting the contract text into an entity naming identification model to extract an element set in the contract text;
step S20, the contract text is input into a reading understanding model to obtain an examination clause analysis result, and a contract examination result is obtained according to a preset expert knowledge base, the template difference result, the element set and the examination clause analysis result.
The contract examination method is applied to contract examination equipment of a financial institution, wherein the contract examination equipment can be a terminal or PC equipment, and for convenience of description, the contract examination equipment is taken as an example for description; the contract examination equipment receives the contract text, inputs the contract text into a text matching model, determines the type of the contract text through the text matching model, determines a standard contract template according to the type of the contract text, and compares the contract text with the standard contract template through the text matching model to obtain a template difference result; the contract examination equipment inputs the contract text into the entity naming recognition model, recognizes and classifies each element in the contract text through the entity naming recognition model, determines the position information corresponding to each element, and extracts the element set in the contract text according to the position information. The contract examination equipment inputs the contract text into the reading understanding model, determines the type of the contract text through the reading understanding model, determines examination terms according to the type of the contract text, identifies the contract text through the reading understanding model to obtain a logic problem, and obtains an analysis result of the examination terms according to the logic problem and the examination terms; the contract examination equipment determines non-compliant elements in the element set according to a preset expert knowledge base, a template difference result, the element set and an examination term analysis result, determines examination opinions corresponding to the non-compliant elements, determines position information of the non-compliant elements in a contract text, and annotates the same text according to the position information and the examination opinions to obtain a contract examination result. It should be noted that the text matching model, the entity naming recognition model and the reading understanding model are obtained by performing model training on different sub-models of a Bert-based model based on corpus training in the financial field.
The contract examination method of the embodiment comprises the steps of receiving a contract text, inputting the contract text into a text matching model to obtain a template difference result, and inputting the contract text into an entity naming recognition model to extract an element set in the contract text; and inputting the contract text into the reading understanding model to obtain an examination clause analysis result, and obtaining a contract examination result according to a preset expert knowledge base, a template difference result, an element set and an examination clause analysis result. According to the method, the contract text is processed through the text matching model, the entity naming recognition model and the reading understanding model, and the contract examination result is obtained by combining the preset expert knowledge base, so that the efficiency and the accuracy of the contract examination are improved.
The respective steps will be described in detail below:
step S10, receiving a contract text, inputting the contract text into a text matching model to obtain a template difference result, and inputting the contract text into an entity naming identification model to extract an element set in the contract text;
in this embodiment, the contract examination device receives the contract text, parses the contract text, inputs the parsed contract text into the text matching model and the entity naming identification model, respectively, processes the contract text through the text matching model to obtain a template difference result, processes the contract text through the entity naming identification model, and extracts an element set in the contract text. It should be noted that the contract text received by the contract inspection apparatus may be in a PDF format or a picture format, where the contract inspection apparatus needs to convert the contract text in the PDF format or the picture format into a document format for the subsequent steps to be performed, and the contract inspection apparatus can also identify the type of the contract text, where the type of the contract text includes but is not limited to: buy and sell contracts, gift contracts, borrow contracts, lease contracts, etc.
Specifically, the step of inputting the contract text into a text matching model to obtain a template difference result includes:
step a, inputting the contract text into a text matching model, determining the type of the contract text through the text matching model, and determining a standard contract template according to the type of the contract text;
in the step, when the contract examination equipment receives the contract text, the contract text is input into a text matching model, the type of the contract text is determined through the text matching model, and a corresponding standard contract template is obtained in a standard contract template database according to the type of the contract text; it should be noted that different types of contracts correspond to different standard contract templates, and the relevant business personnel store all types of standard contract templates in the standard contract template database.
And b, comparing the contract text with the standard contract template through the text matching model to obtain a template difference result.
In the step, after acquiring the standard contract template, the contract examination device compares the contract text with the standard contract template through a text matching model, and marks the part of the contract text which has a difference with the standard contract template to obtain a template difference result; such as: the currency corresponding to the amount of money is not mentioned in the contract text, the currency corresponding to the amount of money is recorded in the standard contract template, and at the moment, the amount of money in the contract text is marked by the text matching model.
Specifically, the step of inputting the contract text into an entity naming recognition model to extract the element set in the contract text comprises the following steps:
and c, inputting the contract text into an entity naming recognition model, recognizing and classifying each element in the contract text through the entity naming recognition model, determining the position information corresponding to each element, and extracting an element set in the contract text according to the position information.
In the step, the contract examination equipment inputs the contract text into an entity naming recognition model, identifies and classifies each element in the contract text through the entity naming recognition model, determines the corresponding position information of each element in the contract text, and extracts an element set in the contract text according to the position information; such as: the contract text comprises but is not limited to characters representing currency, amount, date and the like, the entity naming recognition model recognizes each character in the contract text, further obtains the meaning expressed by each character, classifies the characters into elements such as currency, amount, date and the like according to the meaning represented by each character, then determines the corresponding position information of each element in the contract text, extracts the corresponding elements in the contract text according to the position information, and obtains an element set.
Step S20, the contract text is input into a reading understanding model to obtain an examination clause analysis result, and a contract examination result is obtained according to a preset expert knowledge base, the template difference result, the element set and the examination clause analysis result.
In this embodiment, the contract examination device inputs the contract text into the reading understanding model, determines examination terms corresponding to the contract text through the reading understanding model, obtains an examination term analysis result based on the examination terms, and obtains a contract examination result according to a preset expert knowledge base, a template difference result, an element set and the examination term analysis result. It should be noted that the preset expert knowledge base includes various empirical data of each contract-censored expert during the censoring process.
Specifically, the step of inputting the contract text into the reading understanding model to obtain the examination term analysis result includes:
step d, inputting the contract text into a reading understanding model, determining the type of the contract text through the reading understanding model, and determining an examination clause according to the type of the contract text;
in this step, the contract examination device inputs the contract text into the reading understanding model, determines the type of the contract text through the reading understanding model, and determines the corresponding examination terms in the examination term library according to the type of the contract text, it can be understood that different key elements are included in different types of contract texts, so the examination terms corresponding to different types of contract texts are different, for example, the key elements in the buying and selling contract may include buyer and seller, product for buying and selling, amount of money for buying and selling, etc., the key elements in the borrowing contract may include repayment term, borrower, amount of borrowing, etc., and thus the examination terms corresponding to the type of the contract text need to be determined in the examination term library including the examination terms of different types of contracts.
And e, identifying the contract text through the reading understanding model to obtain a logic problem, and obtaining an analysis result of the examination clause according to the logic problem and the examination clause.
In this step, the contract examination device identifies the contract text by reading the understanding model to obtain a logic problem, and obtains an examination clause analysis result according to the logic problem and the examination clauses. For example, the contract examination device recognizes that the single-use limit of the credit card transacted by the valuable parties in the contract text does not exceed 1 ten thousand yuan of RMB through reading the understanding model, and the reading of the understanding model can obtain the corresponding logic problem, namely the specific name of the valuable parties? "," single use amount and currency? And the like, and according to the logic problem, corresponding logic elements of 'a noble party is a certain precious party', 'the single-use limit does not exceed 1 ten thousand yuan of RMB' and the like are extracted from the contract text, and according to the logic elements, the examination clause of the single-use limit of the credit card corresponding to examination is determined in the examination clause, and the analysis result of the examination clause is obtained based on the logic problem and the examination clause.
Specifically, the step of obtaining the contract examination result according to the preset expert knowledge base, the template difference result, the element set and the examination term analysis result includes:
f, determining non-compliant elements in the element set according to a preset expert knowledge base, the template difference result, the element set and the examination term analysis result, and determining examination opinions corresponding to the non-compliant elements;
in the step, the contract examination equipment determines non-compliant elements in the element set according to a preset expert knowledge base, a template difference result, an element set and an examination term analysis result, and determines examination opinions corresponding to the non-compliant elements; such as: the element set comprises elements such as contract types, currencies, amounts, dates and validity periods, the contract examination equipment determines a corresponding preset expert knowledge base according to the contract types, integrates the template difference result, the element set and the examination term analysis result according to the corresponding preset expert knowledge base, examines whether the semantics of each element in the element set are in compliance, for example, the currency in the contract text is U.S. dollar, and specifies that the contract text of the type only supports RMB according to the preset expert knowledge base, the template difference result and the examination term analysis result, wherein the currency element is a non-compliance element; and after the contract examination equipment determines the non-compliance element, determining the corresponding examination opinion according to a preset expert knowledge base, a template difference result and an examination term analysis result, wherein if the non-compliance element is currency, the corresponding examination opinion is 'changing the currency into RMB', and if the non-compliance element is money, the corresponding examination opinion is 'changing the money into a certain element', and the like.
And g, determining the position information of the non-compliance element in the contract text, and annotating the contract text according to the position information and the examination opinions to obtain a contract examination result.
In the step, when the contract examination equipment determines examination opinions corresponding to the non-compliant elements, the position information of the non-compliant elements in the contract text is determined, and the contract text is annotated according to the position information and the examination opinions; such as: the contract examination equipment determines that the position information of the non-compliant elements in the contract text is the third row, the fifth character to the fourth row, the second character, the comment of the corresponding examination opinions on the corresponding position to complete the comment of the contract text, and the final obtained contract examination result is the contract text with the comment.
The contract examination equipment of the embodiment receives the contract text, inputs the contract text into the text matching model, determines the type of the contract text through the text matching model, determines the standard contract template according to the type of the contract text, and compares the contract text with the standard contract template through the text matching model to obtain a template difference result; the contract examination equipment inputs the contract text into the entity naming recognition model, recognizes and classifies each element in the contract text through the entity naming recognition model, determines the position information corresponding to each element, and extracts the element set in the contract text according to the position information. The contract examination equipment inputs the contract text into the reading understanding model, determines the type of the contract text through the reading understanding model, determines examination terms according to the type of the contract text, identifies the contract text through the reading understanding model to obtain a logic problem, and obtains an analysis result of the examination terms according to the logic problem and the examination terms; the contract examination equipment determines non-compliant elements in the element set according to a preset expert knowledge base, a template difference result, an element set and an examination term analysis result, determines examination opinions corresponding to the non-compliant elements, determines position information of the non-compliant elements in a contract text, and annotates the same text according to the position information and the examination opinions to obtain a contract examination result; the efficiency and the accuracy rate of contract examination are improved.
Further, based on the first embodiment of the contract inspection method of the present invention, a second embodiment of the contract inspection method of the present invention is proposed.
The second embodiment of the treaty examination method differs from the first embodiment of the treaty examination method in that step S10 is preceded by further comprising:
step h, acquiring historical contract examination data, and training different sub-models in a pre-training model according to the historical contract examination data to respectively obtain a text matching pre-model, an entity naming recognition pre-model and a reading understanding pre-model;
and i, acquiring verification contract inspection data, respectively inputting the verification contract inspection data into the text matching pre-model, the entity naming identification pre-model and the reading understanding pre-model to obtain a verification result, and determining the text matching model, the entity naming identification model and the reading understanding model based on the verification result.
In the embodiment, the contract examination equipment acquires historical contract examination data, determines a historical template difference result, a historical element set and a historical examination clause analysis result according to the historical contract examination data, and trains different sub-models in a pre-training model according to the historical template difference result, the historical element set and the historical examination clause analysis result to respectively obtain a text matching pre-model, an entity naming recognition pre-model and a reading understanding pre-model; the contract examination equipment acquires the examination data of the examination contract, respectively inputs the examination data of the examination contract into the text matching pre-model, the entity naming identification pre-model and the reading understanding pre-model to obtain an examination result, and determines the text matching model, the entity naming identification model and the reading understanding model based on the examination result.
The following describes the steps in detail:
step h, acquiring historical contract examination data, and training different sub-models in a pre-training model according to the historical contract examination data to respectively obtain a text matching pre-model, an entity naming recognition pre-model and a reading understanding pre-model;
in the step, the contract examination equipment acquires historical contract examination data, determines a historical template difference result, a historical element set and a historical examination term analysis result according to the historical contract examination data, trains a sub-model in a pre-training model according to the historical template difference result to obtain a text matching pre-model, trains the sub-model in the pre-training model according to the historical element set to obtain an entity naming recognition pre-model, and trains the sub-model in the pre-training model according to the historical examination term analysis result to obtain a reading understanding pre-model. It should be noted that the pre-training model includes, but is not limited to: bert, eletra, roberta and the like, and the efficiency of model training and the accuracy of model identification can be accelerated by training the pre-trained model; the historical contract inspection data is the data of contracts which pass through manual inspection at historical time, and comprises historical template difference results, historical element sets and historical inspection clause analysis results corresponding to different types of contracts.
And i, acquiring verification contract inspection data, respectively inputting the verification contract inspection data into the text matching pre-model, the entity naming identification pre-model and the reading understanding pre-model to obtain a verification result, and determining the text matching model, the entity naming identification model and the reading understanding model based on the verification result.
In this step, the contract examination device obtains the verification contract examination data, and respectively inputs the verification contract examination data into the text matching pre-model, the entity naming identification pre-model and the reading understanding pre-model to obtain the verification result, and determines the text matching model, the entity naming identification model and the reading understanding model based on the verification result, such as: the contract examination equipment randomly selects part of the historical contract examination data in the historical contract examination data as verification contract examination data, inputs a verification template difference result in the verification contract examination data into the text matching pre-model to obtain a first identification accuracy after training to obtain a text matching pre-model, an entity naming recognition pre-model and a reading understanding pre-model, inputs a verification element set in the verification contract examination data into the entity naming recognition pre-model to obtain a second identification accuracy, inputs a verification examination clause analysis result in the verification contract examination data into the reading understanding pre-model to obtain a third identification accuracy, respectively compares the first identification accuracy, the second identification accuracy and the third identification accuracy with a preset accuracy threshold, and trains the text matching pre-model, the naming entity recognition pre-model and/or the reading understanding pre-model again if the identification accuracy is less than the preset accuracy threshold, and if the recognition accuracy is not less than the preset accuracy threshold, determining that the text matching pre-model is used as a final text matching model, determining that the entity naming recognition pre-model is used as a final entity naming recognition model, and determining that the reading understanding pre-model is used as a final reading understanding model.
The contract examination equipment of the embodiment acquires historical contract examination data, determines a historical template difference result, a historical element set and a historical examination term analysis result according to the historical contract examination data, and trains different sub-models in a pre-training model according to the historical template difference result, the historical element set and the historical examination term analysis result to respectively obtain a text matching pre-model, an entity naming recognition pre-model and a reading understanding pre-model; the contract examination equipment acquires the examination data of the examination contract, respectively inputs the examination data of the examination contract into the text matching pre-model, the entity naming identification pre-model and the reading understanding pre-model to obtain an examination result, and determines the text matching model, the entity naming identification model and the reading understanding model based on the examination result, which is beneficial to improving the efficiency and the accuracy of the contract examination.
Further, a third embodiment of the contract inspection method of the present invention is proposed based on the first and second embodiments of the contract inspection method of the present invention.
The third embodiment of the treaty examination method differs from the first and second embodiments of the treaty examination method in that step S20 is followed by further comprising:
and j, updating historical contract examination data according to the contract examination result, and performing optimization operation on the text matching model, the entity naming recognition model and the reading understanding model according to the updated historical contract examination data.
In this embodiment, after the contract examination device examines the contract texts, the contract examination device updates the historical contract examination data according to the contract examination result, and performs optimization operation on the text matching model, the entity naming recognition model and the reading understanding model according to the updated historical contract examination data. Not only provides abundant linguistic data for model optimization, but also deposits relevant business knowledge, and further improves the accuracy of contract examination.
The present invention also provides a contract examination apparatus, including:
the extraction module is used for receiving a contract text, inputting the contract text into a text matching model to obtain a template difference result, and inputting the contract text into an entity naming identification model to extract an element set in the contract text;
and the examination module is used for inputting the contract text into a reading understanding model to obtain an examination term analysis result and obtaining a contract examination result according to a preset expert knowledge base, the template difference result, the element set and the examination term analysis result.
Preferably, the extraction module is further configured to:
inputting the contract text into a text matching model, determining the type of the contract text through the text matching model, and determining a standard contract template according to the type of the contract text;
and comparing the contract text with the standard contract template through the text matching model to obtain a template difference result.
Preferably, the extraction module is further configured to:
inputting the contract text into an entity naming recognition model, recognizing and classifying each element in the contract text through the entity naming recognition model, determining position information corresponding to each element, and extracting an element set in the contract text according to the position information.
Preferably, the extraction module is further configured to:
inputting the contract text into a reading understanding model, determining the type of the contract text through the reading understanding model, and determining examination terms according to the type of the contract text;
and identifying the contract text through the reading understanding model to obtain a logic problem, and obtaining an analysis result of the examination clause according to the logic problem and the examination clause.
Preferably, the extraction module is further configured to:
determining non-compliant elements in the element set according to a preset expert knowledge base, the template difference result, the element set and the examination term analysis result, and determining examination opinions corresponding to the non-compliant elements;
and determining the position information of the non-compliance element in the contract text, and annotating the contract text according to the position information and the examination opinions to obtain a contract examination result.
Preferably, the extraction module further comprises a training module configured to:
acquiring historical contract examination data, and training different sub-models in a pre-training model according to the historical contract examination data to respectively obtain a text matching pre-model, an entity naming recognition pre-model and a reading understanding pre-model;
and acquiring verification contract inspection data, inputting the verification contract inspection data into the text matching pre-model, the entity naming identification pre-model and the reading understanding pre-model respectively to obtain verification results, and determining the text matching model, the entity naming identification model and the reading understanding model based on the verification results.
Preferably, the review module further comprises an update module configured to:
and updating historical contract examination data according to the contract examination result, and performing optimization operation on the text matching model, the entity naming identification model and the reading understanding model according to the updated historical contract examination data.
The invention also provides a contract examination system.
The contract examination system of the invention comprises: a memory, a processor, and a treaty examination program stored on the memory and executable on the processor, the treaty examination program when executed by the processor implementing the steps of the treaty examination method as described above.
The method implemented when the contract inspection program executed on the processor is executed may refer to each embodiment of the contract inspection method of the present invention, and is not described herein again.
The invention also provides a computer readable storage medium.
The computer-readable storage medium of the present invention has stored thereon a treaty review program that, when executed by a processor, implements the steps of the treaty review method described above.
The method implemented when the contract inspection program executed on the processor is executed may refer to each embodiment of the contract inspection method of the present invention, and is not described herein again.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium (e.g., ROM/RAM, magnetic disk, optical disk) as described above and includes instructions for enabling a terminal device (e.g., a mobile phone, a computer, a server, or a network device) to execute the method according to the embodiments of the present invention.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (10)
1. A method of treaty review, comprising the steps of:
receiving a contract text, inputting the contract text into a text matching model to obtain a template difference result, and inputting the contract text into an entity naming identification model to extract an element set in the contract text;
and inputting the contract text into a reading understanding model to obtain an examination term analysis result, and obtaining a contract examination result according to a preset expert knowledge base, the template difference result, the element set and the examination term analysis result.
2. The method of treaty review of claim 1, wherein the step of entering the treaty text into a text matching model to obtain a template difference result comprises:
inputting the contract text into a text matching model, determining the type of the contract text through the text matching model, and determining a standard contract template according to the type of the contract text;
and comparing the contract text with the standard contract template through the text matching model to obtain a template difference result.
3. The method of treaty review of claim 1, wherein the step of entering the treaty text into an entity naming recognition model to extract a set of elements in the treaty text comprises:
inputting the contract text into an entity naming recognition model, recognizing and classifying each element in the contract text through the entity naming recognition model, determining position information corresponding to each element, and extracting an element set in the contract text according to the position information.
4. The method of treaty review as claimed in claim 1, wherein said step of entering said treaty text into a reading understanding model to obtain review term analysis results comprises:
inputting the contract text into a reading understanding model, determining the type of the contract text through the reading understanding model, and determining examination terms according to the type of the contract text;
and identifying the contract text through the reading understanding model to obtain a logic problem, and obtaining an analysis result of the examination clause according to the logic problem and the examination clause.
5. The method of treaty review according to claim 1, wherein said step of obtaining treaty review results based on a pre-defined expert knowledge base, said template difference results, said set of elements, and said review term analysis results comprises:
determining non-compliant elements in the element set according to a preset expert knowledge base, the template difference result, the element set and the examination term analysis result, and determining examination opinions corresponding to the non-compliant elements;
and determining the position information of the non-compliance element in the contract text, and annotating the contract text according to the position information and the examination opinions to obtain a contract examination result.
6. The method of treaty review of claim 1, wherein, prior to the step of receiving treaty text, the method of treaty review comprises:
acquiring historical contract examination data, and training different sub-models in a pre-training model according to the historical contract examination data to respectively obtain a text matching pre-model, an entity naming recognition pre-model and a reading understanding pre-model;
and acquiring verification contract inspection data, inputting the verification contract inspection data into the text matching pre-model, the entity naming identification pre-model and the reading understanding pre-model respectively to obtain verification results, and determining the text matching model, the entity naming identification model and the reading understanding model based on the verification results.
7. The method of treaty review according to claim 1, wherein after the step of obtaining a treaty review result based on a pre-set expert knowledge base, the template difference results, the set of elements, and the review term analysis results, the method of treaty review comprises:
and updating historical contract examination data according to the contract examination result, and performing optimization operation on the text matching model, the entity naming identification model and the reading understanding model according to the updated historical contract examination data.
8. A treaty examination apparatus, characterized in that the treaty examination apparatus comprises:
the extraction module is used for receiving a contract text, inputting the contract text into a text matching model to obtain a template difference result, and inputting the contract text into an entity naming identification model to extract an element set in the contract text;
and the examination module is used for inputting the contract text into a reading understanding model to obtain an examination term analysis result and obtaining a contract examination result according to a preset expert knowledge base, the template difference result, the element set and the examination term analysis result.
9. A system for treaty review, the system comprising: a memory, a processor, and a treaty examination program stored on the memory and executable on the processor, the treaty examination program when executed by the processor implementing the steps of the treaty examination method of any of claims 1-7.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a treaty examination program, which when executed by a processor implements the steps of the treaty examination method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210165386.6A CN114549241A (en) | 2022-02-22 | 2022-02-22 | Contract examination method, device, system and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210165386.6A CN114549241A (en) | 2022-02-22 | 2022-02-22 | Contract examination method, device, system and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114549241A true CN114549241A (en) | 2022-05-27 |
Family
ID=81676934
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210165386.6A Pending CN114549241A (en) | 2022-02-22 | 2022-02-22 | Contract examination method, device, system and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114549241A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116384387A (en) * | 2023-01-04 | 2023-07-04 | 深圳擎盾信息科技有限公司 | Automatic combination and examination method and device |
| CN117011048A (en) * | 2023-07-25 | 2023-11-07 | 明亚保险经纪股份有限公司 | Online entrusting protocol signing method, system, equipment and medium |
| CN117151096A (en) * | 2023-09-05 | 2023-12-01 | 江苏群杰物联科技有限公司 | Intelligent contract checking method and device, electronic equipment and storage medium |
| CN117172249A (en) * | 2023-11-03 | 2023-12-05 | 青矩技术股份有限公司 | Contract checking method, device, equipment and computer readable storage medium |
-
2022
- 2022-02-22 CN CN202210165386.6A patent/CN114549241A/en active Pending
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116384387A (en) * | 2023-01-04 | 2023-07-04 | 深圳擎盾信息科技有限公司 | Automatic combination and examination method and device |
| CN117011048A (en) * | 2023-07-25 | 2023-11-07 | 明亚保险经纪股份有限公司 | Online entrusting protocol signing method, system, equipment and medium |
| CN117011048B (en) * | 2023-07-25 | 2024-03-26 | 明亚保险经纪股份有限公司 | Online entrusting protocol signing method, system, equipment and medium |
| CN117151096A (en) * | 2023-09-05 | 2023-12-01 | 江苏群杰物联科技有限公司 | Intelligent contract checking method and device, electronic equipment and storage medium |
| CN117172249A (en) * | 2023-11-03 | 2023-12-05 | 青矩技术股份有限公司 | Contract checking method, device, equipment and computer readable storage medium |
| CN117172249B (en) * | 2023-11-03 | 2024-01-26 | 青矩技术股份有限公司 | Contract checking method, device, equipment and computer readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111177569B (en) | Recommendation processing method, device and equipment based on artificial intelligence | |
| CN114549241A (en) | Contract examination method, device, system and computer readable storage medium | |
| CN110597964B (en) | Double-recording quality inspection semantic analysis method and device and double-recording quality inspection system | |
| US20220292861A1 (en) | Docket Analysis Methods and Systems | |
| CN112434501A (en) | Work order intelligent generation method and device, electronic equipment and medium | |
| CN107341157B (en) | Customer service conversation clustering method and device | |
| CN113935710A (en) | Contract auditing method and device, electronic equipment and storage medium | |
| CN114626731A (en) | Risk identification method and device, electronic equipment and computer readable storage medium | |
| CN111782793A (en) | Intelligent customer service processing method, system and equipment | |
| CN115357699A (en) | Text extraction method, device, equipment and storage medium | |
| CN114219507A (en) | Qualification auditing method and device for traditional Chinese medicine supplier, electronic equipment and storage medium | |
| CN117520503A (en) | Financial customer service dialogue generation method, device, equipment and medium based on LLM model | |
| CN114549177A (en) | Insurance letter examination method, device, system and computer readable storage medium | |
| CN114998920B (en) | Supply chain financial file management method and system based on NLP semantic recognition | |
| CN111401854A (en) | Information processing method and device | |
| CN116071150A (en) | Data processing method, bank product popularization, wind control system, server and medium | |
| CN112328812B (en) | Domain knowledge extraction method and system based on self-adjusting parameters and electronic equipment | |
| CN114444040A (en) | Authentication processing method, authentication processing device, storage medium and electronic equipment | |
| CN114741501A (en) | Public opinion early warning method and device, readable storage medium and electronic equipment | |
| CN113886538A (en) | Medical insurance reimbursement information query method and device, electronic equipment and storage medium | |
| CN114067343A (en) | Data set construction method, model training method and corresponding device | |
| CN111144409A (en) | Order following, accepting and examining processing method and system | |
| CN111783432A (en) | Generation method and device of credit certificate examination key point list | |
| CN110232328A (en) | A kind of reference report analytic method, device and computer readable storage medium | |
| CN116092496B (en) | Method and system for judging speech expression integrity |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |