Data verification method, device, computer equipment and storage medium

Abstract

The present invention relates to the field of artificial intelligence technology, and discloses a data verification method, device, computer equipment, and storage medium, including: detecting an instruction to report an agricultural insurance clause document, obtaining the agricultural insurance clause document to be reported; identifying the compensation standard of the agricultural insurance clause document to be reported, converting the identified compensation standard into structured data, and recording it as a target compensation standard; storing the target compensation standard in a preset agricultural insurance compensation standard database, calculating the semantic similarity between the target compensation standard and several original agricultural insurance compensation standards in the agricultural insurance compensation standard database, and verifying the duplication between the target compensation standard and the several original agricultural insurance compensation standards based on the semantic similarity; if duplication exists, generating an alarm message. The embodiment of the present invention can automate processing, reduce clause review time, save labor costs by reducing the workload of manual review, reduce conflicts caused by duplication of clauses, and save operating costs.

Description

Technical Field

[0001] The present invention relates to the fields of artificial intelligence technology and financial technology, and in particular to a data verification method, device, computer equipment and storage medium. Background Art

[0002] In the field of agricultural insurance (abbreviated as agricultural insurance), insurance clauses are typically customized based on the different insured objects (such as crops, livestock, etc.) and different regions. This customized clause design requires consideration of numerous factors, such as climate, geographical environment, and agricultural production methods. As a result, agricultural insurance clauses are highly heterogeneous. As the business continues to expand and change, agricultural insurance clauses frequently need to be updated and adjusted to adapt to new market demands and environmental changes. The introduction of new clauses often leads to duplication of compensation scope and compensation ratios, which in turn increases management costs and complexity for insurance companies and may lead to compensation disputes.

[0003] Agricultural insurance clauses have distinct characteristics distinct from other types of insurance. They require high levels of regionality, specificity, and customization, making it impossible to cover all underlying assets with a single, unified policy. This has led to a large volume of agricultural insurance clause filings and a high backlog. Over the past two years, over 1,000 agricultural insurance clause filings have been filed. In this business context, institutions may develop entirely new insurance clauses for new targets, inevitably leading to duplicate insurance liabilities and overlapping compensation coverages. This leads to duplication and homogeneity in the filing process. Currently, clause verification requires manual verification, with an average processing time of 11 hours per clause. Cross-checking different clauses further increases the time required. Therefore, the repetitive verification of clause liabilities remains a time-consuming and labor-intensive pain point in the business process, while increasing manpower is prohibitively expensive and fails to deliver corresponding value.

[0004] At present, the scope of compensation and the ratio of compensation in agricultural insurance clauses require manual data verification during the updating and comparison process of the clauses. Manual comparison of the contents of these clauses is time-consuming and prone to errors.

[0005] Therefore, the existing technology still needs to be improved and developed. Summary of the Invention

[0006] In view of the above-mentioned deficiencies in the prior art, the present invention provides a data verification method, device, computer equipment and storage medium, which aims to solve the problem in the prior art that the scope of compensation and the compensation ratio in agricultural insurance clauses require manual data verification during the updating and comparison process of the clauses, and manual comparison of the contents of these clauses is time-consuming and prone to errors.

[0007] The technical solutions of the present invention are as follows:

[0008] A first embodiment of the present invention provides a data verification method, the method comprising:

[0009] Detects the agricultural insurance clause document filing instruction and obtains the agricultural insurance clause document to be filed;

[0010] Identifying the compensation standard of the agricultural insurance clause document to be reported, and converting the identified compensation standard into structured data, which is recorded as the target compensation standard;

[0011] Storing the target compensation standard in a preset agricultural insurance compensation standard database, wherein the preset agricultural insurance compensation standard database stores a plurality of original agricultural insurance compensation standards;

[0012] Calculating semantic similarities between the target compensation standard and the plurality of original agricultural insurance compensation standards respectively, and verifying duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards based on the semantic similarities;

[0013] If the target compensation standard is repeated with the several original agricultural insurance compensation standards, an alarm message is generated.

[0014] Another embodiment of the present invention provides a data verification device, the device comprising:

[0015] A data acquisition module is used to detect an instruction to report agricultural insurance clause documents and obtain the agricultural insurance clause documents to be reported;

[0016] A data conversion module is used to identify the compensation standard of the agricultural insurance clause document to be reported, and convert the identified compensation standard into structured data, which is recorded as the target compensation standard;

[0017] A data storage module, configured to store the target compensation standard in a preset agricultural insurance compensation standard database, wherein the preset agricultural insurance compensation standard database stores a plurality of original agricultural insurance compensation standards;

[0018] a data verification module, configured to respectively calculate the semantic similarity between the target compensation standard and the plurality of original agricultural insurance compensation standards, and verify the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards based on the semantic similarity;

[0019] The warning information generation module is used to generate a warning message if the target compensation standard is repeated with the several original agricultural insurance compensation standards.

[0020] Another embodiment of the present invention provides a computer device, the computer device comprising at least one processor; and

[0021] a memory communicatively connected to the at least one processor; wherein,

[0022] The memory stores instructions that can be executed by the at least one processor. The instructions are executed by the at least one processor to enable the at least one processor to perform the steps of the above-mentioned data verification method.

[0023] Another embodiment of the present invention further provides a computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are executed by one or more processors, the one or more processors can perform the steps of the above-mentioned data verification method.

[0024] Beneficial effect: The data verification method of the embodiment of the present invention can obtain the agricultural insurance clause document to be reported by detecting the agricultural insurance clause document reporting instruction; identify the compensation standard of the agricultural insurance clause document to be reported, and convert the identified compensation standard into structured data, which is recorded as the target compensation standard; store the target compensation standard in a preset agricultural insurance compensation standard database, wherein the preset agricultural insurance compensation standard database stores a number of original agricultural insurance compensation standards; calculate the semantic similarity between the target compensation standard and the several original agricultural insurance compensation standards respectively, and verify the duplication between the target compensation standard and the several original agricultural insurance compensation standards according to the semantic similarity; if there is duplication between the target compensation standard and the several original agricultural insurance compensation standards, an alarm message is generated. In the present invention, the automated processing of extracting and comparing the compensation standards of agricultural insurance clause documents reduces the clause review time by more than 80%, from an average of 11 hours for each clause review to less than 20 minutes; the automated semantic similarity calculation reduces the workload of manual review and saves labor costs; when there are duplications in agricultural insurance compensation standards, an alarm message is generated, and staff can modify the agricultural insurance clauses based on the alarm message, thereby reducing duplications and conflicts in clauses and saving operating costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0026] Figure 1 A schematic diagram of an application environment of an embodiment of a data verification method of the present invention;

[0027] Figure 2 A flow chart of a preferred embodiment of a data verification method of the present invention;

[0028] Figure 3 for Figure 2 A flowchart of a specific embodiment of step S200;

[0029] Figure 4 for Figure 2 A flowchart of a specific embodiment of step S300;

[0030] Figure 5 for Figure 2 A flowchart of a specific embodiment of step S400;

[0031] Figure 6 for Figure 5 A flowchart of a specific embodiment of step S402;

[0032] Figure 7 A schematic diagram of functional modules of a preferred embodiment of a data verification device of the present invention;

[0033] Figure 8 A schematic structural diagram of a preferred embodiment of a computer device of the present invention;

[0034] Figure 9 This is another structural diagram of a preferred embodiment of a computer device of the present invention. DETAILED DESCRIPTION

[0035] To make the purpose, technical solution and effect of the present invention clearer and more specific, the present invention is further described in detail below. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0036] The following describes the embodiments of the present invention in conjunction with the accompanying drawings. Those skilled in the art will appreciate that, with the development of technology and the emergence of new scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0037] The terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the terms used in this way can be interchangeable under appropriate circumstances, and this is merely a way of distinguishing the objects of the same attributes when describing them in the embodiments of the present application. Here, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, so that the process, method, system, product or equipment comprising a series of units need not be limited to those units, but may include other units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0038] The method provided in this application can be applied to artificial intelligence (AI) scenarios. AI is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a branch of computer science that attempts to understand the essence of intelligence and produce a new type of intelligent machine that can respond in a similar way to human intelligence. Artificial intelligence is also the study of the design principles and implementation methods of various intelligent machines, so that machines have the functions of perception, reasoning and decision-making. Research in the field of artificial intelligence includes robotics, natural speech processing, computer vision, decision-making and reasoning, human-computer interaction, recommendation and search, basic AI theory, etc.

[0039] The intelligent question-answering processing method based on artificial intelligence provided by the embodiment of the present invention can be applied in Figure 1 In an application environment, the client communicates with the server through a network. The server can detect the agricultural insurance clause document reporting instruction through the client and obtain the agricultural insurance clause document to be reported; identify the compensation standard of the agricultural insurance clause document to be reported, convert the identified compensation standard into structured data, and record it as the target compensation standard; store the target compensation standard in a preset agricultural insurance compensation standard database, wherein the preset agricultural insurance compensation standard database stores several original agricultural insurance compensation standards; calculate the semantic similarity between the target compensation standard and the several original agricultural insurance compensation standards respectively, and sort out the duplication between the target compensation standard and the several original agricultural insurance compensation standards according to the semantic similarity. Verification is performed; if the target compensation standard is repeated with the several original agricultural insurance compensation standards, an alarm message is generated. In the present invention, the extraction and comparison of the compensation standards of the agricultural insurance clause documents are automated, which reduces the clause review time by more than 80%, from an average of 11 hours for each clause review to less than 20 minutes; the automated semantic similarity calculation reduces the workload of manual review and saves labor costs; when there is duplication of agricultural insurance compensation standards, an alarm message is generated, and the staff can modify the agricultural insurance clauses through the alarm message, thereby reducing the duplication and conflict of clauses and saving operating costs. Among them, the client can be but is not limited to various personal computers, laptops, smart phones, tablet computers and portable wearable devices. The server can be implemented with an independent server or a server cluster composed of multiple servers. The present invention is described in detail below through specific embodiments.

[0040] In order to solve the above problems, the present invention provides a data verification method. Figure 2 , Figure 2 This is a flow chart of a preferred embodiment of a data verification method of the present invention. Figure 2 As shown, it includes:

[0041] Step S100: Detecting an agricultural insurance clause document filing instruction and obtaining the agricultural insurance clause document to be filed.

[0042] The data verification method in the embodiment of the present invention is applied to a server that verifies the duplication between the compensation standard of the reported agricultural insurance clause and the existing compensation standard in an agricultural insurance scenario. The server can verify the duplication.

[0043] When a new agricultural insurance policy appears at an insurance institution, it must be reported and filed. This filing requires the generation of an agricultural insurance policy document. This document includes, but is not limited to, the policy table data generated by the institution number, underwriting institution, product code, product name, crop growth period, and compensation scope.

[0044] The table data in the agricultural insurance clause document to be reported may be in non-docx format, unclearly described tables, simple tables, or complex tables (such as those with merged cells).

[0045] Step S200: Identify the compensation standard of the agricultural insurance clause document to be reported, convert the identified compensation standard into structured data, and record it as the target compensation standard.

[0046] The compensation standards for agricultural insurance vary depending on factors such as crop type, growing period, degree of loss and regional policies.

[0047] Agricultural insurance includes crop insurance and livestock insurance compensation. For example, an example of the compensation standard for major crop insurance in crop insurance is as follows:

[0048] Corn: The premium is typically 18 yuan per mu (approximately 1.5 acres) with an insured amount of 400 yuan per mu. Compensation is calculated based on the loss rate and the damaged area. When the loss rate reaches a certain percentage (e.g., above 80%), it may be considered a total loss and full compensation will be paid according to the insured amount.

[0049] Wheat: Premium 18 yuan / mu, insured amount 450 yuan / mu. Compensation is the same as for corn.

[0050] Rice: The premium is 30 yuan per mu (approximately 1.5 acres). The insured amount may vary depending on the policy, for example, 650 yuan per mu in some areas and higher (e.g., 1,100 yuan per mu) in others. Compensation is also calculated based on the loss rate and the damaged area, with a minimum payout threshold (e.g., 30%).

[0051] Other crops: such as peanuts, apples, etc., the premiums and insured amounts are also different, and the compensation standards need to be determined according to the specific crops and insurance terms.

[0052] Compensation for livestock insurance (e.g., for breeding sows, dairy cows, and fattening pigs) is typically determined by the insured amount and the extent of the loss. For example, the insured amount for breeding sows might be 1,500 yuan per head. If a loss occurs within the insurance coverage, the compensation amount will be determined based on the extent of the loss and the policy terms.

[0053] Image recognition or artificial intelligence technology is used to identify the compensation standards in the agricultural insurance policy documents to be filed and extract the identified content. This content includes but is not limited to the institution number, underwriting institution, product code, product name, crop growth period, and compensation scope. The extracted compensation standards are converted into structured data and recorded as the target compensation standards.

[0054] Among them, such as Figure 3 As shown, in step S200, the compensation standard of the agricultural insurance clause document to be reported is identified, and the identified compensation standard is converted into structured data, which is recorded as the target compensation standard, including:

[0055] Step S201: Identify the compensation standard of the agricultural insurance clause document to be reported based on the multimodal large model to obtain the identified compensation standard;

[0056] Step S202: convert the identified compensation standard into structured data in a preset format, which is recorded as the target compensation standard.

[0057] Use a multimodal big model to automatically identify and extract table content from agricultural insurance policy documents. Specifically, use optical character recognition (OCR) technology from the multimodal big model to extract text information from scanned or image-based policy documents. Structure the table content into a comparable data format.

[0058] The modal big model is a machine learning technology based on deep learning. Its core concept is to fuse different media data (such as text, images, audio, and video) and achieve more intelligent information processing by learning the relationships between different modalities. Multimodal big models refer to models that combine multiple modal information such as text, images, video, and audio for training.

[0059] Large multimodal models enable: Cross-modal understanding: They can understand not only information from a single modality but also the relationships between them across modalities. Superior performance: They have achieved excellent performance in multiple fields, such as medical imaging diagnosis, intelligent security, and smart transportation. Strong scalability: They can be applied to diverse tasks and provide diverse intelligent services.

[0060] The multimodal large model first learns through modal representation: converting data of different modalities into a unified data representation, such as converting images into feature vectors and text into word vectors.

[0061] Next, feature fusion is performed: representations of different modalities are combined to obtain more comprehensive information. Common fusion methods include concatenation, weighted averaging, and attention mechanisms.

[0062] The attention mechanism is introduced again: it enables the model to focus on the most important information in different modalities, improving the accuracy and efficiency of information processing.

[0063] Finally, multi-task learning; by learning multiple related tasks at the same time, more comprehensive information fusion and more efficient model training can be achieved.

[0064] Optical Character Recognition (OCR) is a technology that converts text in scanned or photographed images into editable, searchable electronic text. OCR technology is widely used in document processing, digital libraries, automated office work, data entry, and accessible reading, among other fields.

[0065] OCR technology usually includes the following steps:

[0066] Image preprocessing: Preprocess the input image, including removing noise, adjusting contrast, binarization, etc., to better recognize text.

[0067] Layout analysis: Perform layout analysis on the pre-processed image to identify text areas, image areas, table areas, etc., and determine the arrangement direction of the text (such as horizontal, vertical, tilted, etc.).

[0068] Character segmentation: Segmenting text in a text area into individual characters or words is one of the key steps in OCR recognition.

[0069] Feature extraction: Extract features of each character or word, including shape, size, strokes, etc., for subsequent classification and recognition.

[0070] Character recognition: Using machine learning or pattern recognition algorithms, the extracted features are matched with predefined character templates or training sets to determine the specific content of each character or word.

[0071] Post-processing: Post-process the recognition results, including error correction, format adjustment, and reformatting, to improve the accuracy and readability of the recognition results.

[0072] OCR can be used to identify the compensation standard table of agricultural insurance clause documents in image format, and the identified data can be organized into structured data according to the specified format and recorded as the target compensation standard.

[0073] The use of a large multimodal model for table extraction and data verification achieved an accuracy rate of over 95%, significantly higher than the average accuracy of 70% for traditional comparisons.

[0074] Step S300: storing the target compensation standard in a preset agricultural insurance compensation standard database, wherein the preset agricultural insurance compensation standard database stores a plurality of original agricultural insurance compensation standards.

[0075] We obtain existing agricultural insurance clauses in advance. Based on the OCR capability of the multimodal large model, we identify and extract the tables in the document. After extraction, we organize the clauses of each institution by institution and store them in the preset agricultural insurance compensation standard database. In this application, when obtaining the target compensation standard, in order to facilitate data comparison, the target compensation standard is stored in the preset agricultural insurance compensation standard database. Identify through the crop insurance clauses and store the extracted fields in the database to form the basis of agricultural insurance data.

[0076] Among them, such as Figure 4 As shown, in step S300, the target compensation standard is stored in a preset agricultural insurance compensation standard database, wherein the preset agricultural insurance compensation standard database stores a number of original agricultural insurance compensation standards, including:

[0077] Step S301: Based on the agricultural insurance clause document to be reported, obtain the issuing institution corresponding to the target compensation standard;

[0078] Step S302: storing the target compensation standard in a preset agricultural insurance compensation standard database according to the issuing institution, wherein the preset agricultural insurance compensation standard database stores a plurality of original agricultural insurance compensation standards.

[0079] Parse the agricultural insurance clause documents to be filed and obtain the issuing agency of the target compensation standard in the agricultural insurance clause documents to be filed. For ease of management, the target compensation standard can be stored in a preset agricultural insurance compensation standard database according to the issuing agency.

[0080] Insurance issuers primarily refer to insurance companies, the core players in the insurance market. They design, issue, and manage various insurance products to disperse and transfer risk and provide financial protection for customers. Insurance institutions are legally established financial institutions that operate insurance businesses. They collect premiums, establish insurance funds, and assume liability for compensating property losses resulting from agreed-upon insurance events. Insurance institutions in this article are categorized by province.

[0081] The present invention extracts the table corresponding to the compensation standard and stores it in the database, which can be used as the basic input business data of the loss estimation project and become the data basis of agricultural insurance.

[0082] Step S400: Calculate the semantic similarity between the target compensation standard and the plurality of original agricultural insurance compensation standards respectively, and verify the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards according to the semantic similarity.

[0083] Natural language processing (NLP) technology is used to perform semantic analysis and comparison on the table contents in different agricultural insurance clauses. By calculating the semantic similarity of the agricultural insurance compensation standards, the duplication of the compensation scope between the agricultural insurance compensation clause to be reported and the existing original agricultural insurance compensation standards is obtained. In an embodiment of the present invention, the duplication of the target compensation standard and several original agricultural insurance compensation standards is verified one by one, wherein the duplication includes but is not limited to the overlap of the compensation scope and compensation ratio in the compensation standard. For example, if the target compensation standard corresponding to the agricultural insurance compensation clause to be reported stipulates that the damage area of ​​wheat during the growing period is greater than or equal to 40%, then compensation is required. If the original agricultural insurance compensation standard of the original agricultural insurance compensation clause stipulates that the damage area of ​​wheat during the growing period is greater than or equal to 30%, then compensation is required. At this time, it is determined that there is a duplication of the compensation scope between the agricultural insurance compensation clause to be reported and the existing original agricultural insurance compensation standard. Among them, NLP, the full name of Natural Language Processing, is an important direction in the fields of computer science and artificial intelligence. NLP studies the various theories and methods for effective communication between humans and computers using natural language. It integrates linguistics, computer science, and mathematics. Therefore, research in this field involves natural language—the language we use everyday—and is closely linked to linguistics. Natural language processing technologies typically encompass text processing, semantic understanding, discourse analysis, machine translation, information extraction, question-and-answer systems, dialogue systems, text classification, and sentiment analysis. Natural language processing can reduce manual review workload and save labor costs.

[0084] Among them, such as Figure 5 As shown, step S400, i.e., respectively calculating the semantic similarity between the target compensation standard and the plurality of original agricultural insurance compensation standards, and verifying the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards based on the semantic similarity, includes:

[0085] Step S401: input the target compensation standard and the plurality of original agricultural insurance compensation standards into a semantic similarity calculation model;

[0086] Step S402: Obtain the semantic similarity output by the semantic similarity calculation model, and verify the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards based on the semantic similarity.

[0087] The target compensation standard and the multiple original agricultural insurance compensation standards are input into a semantic similarity calculation model, which is an artificial intelligence model. The semantic similarity calculation model is obtained by pre-collecting annotated samples of agricultural insurance compensation standard data and training the artificial intelligence model based on the annotated samples. The semantic similarity calculation model then obtains a semantic similarity value output by the semantic similarity calculation model. Based on the semantic similarity value, the target compensation standard and the multiple original agricultural insurance compensation scopes and compensation ratios are verified for overlap.

[0088] Among them, such as Figure 6 As shown, step S402, i.e., obtaining the semantic similarity output by the semantic similarity calculation model, and verifying the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards according to the semantic similarity, includes:

[0089] Step S421: Obtaining the semantic similarity output by the semantic similarity calculation model, wherein the semantic similarity calculation model is a deep learning model;

[0090] Step S422: determine whether the semantic similarity is greater than or equal to a preset similarity threshold; if so, execute step S423; if not, execute step S424;

[0091] Step S423: determining whether the target compensation standard overlaps with the plurality of original agricultural insurance compensation standards;

[0092] Step S424: determine whether the target compensation standard and the plurality of original agricultural insurance compensation standards are duplicated.

[0093] The semantic similarity calculation model uses a deep learning model. The deep learning model is a machine learning model based on artificial neural networks (ANN). Its core idea is to learn the feature representation and pattern of data through a multi-level neural network structure. The deep learning model includes:

[0094] Multi-layered structure: Deep learning models consist of multiple layers, including an input layer, multiple hidden layers, and an output layer. Each layer contains multiple neurons, each connected to all neurons in the previous layer. This multi-layered structure enables the model to gradually learn abstract features of the data.

[0095] Feature Representation Learning: Deep learning models achieve pattern recognition and data modeling by learning feature representations of data in multiple hidden layers. Each layer can be seen as an abstract representation of the data, and higher-level representations contain more abstract features of the data.

[0096] End-to-end learning: Deep learning models are typically end-to-end learning models, meaning they learn directly from raw input data to the final output without manually designing feature extractors or intermediate representations. This makes the model more flexible and powerful.

[0097] Backpropagation algorithm: Deep learning models are typically trained using the backpropagation algorithm. This algorithm calculates the gradient of the loss function with respect to the model parameters and then uses an optimization algorithm such as gradient descent to update the model parameters to minimize the loss function.

[0098] Using a deep learning model, the semantic similarity of the table content is calculated to determine whether it is greater than or equal to a preset similarity threshold. The preset similarity threshold can be set according to the situation. If the semantic similarity is greater than or equal to the preset similarity threshold, it is determined that the target compensation standard and the several original agricultural insurance compensation standards are duplicated. Otherwise, there is no duplication. This allows for the detection of duplications in compensation scope and compensation ratio.

[0099] Step S500: If the target compensation standard is repeated with the several original agricultural insurance compensation standards, an alarm message is generated.

[0100] If the target compensation standard is repeated with any original agricultural insurance compensation standard in the database, an alarm message will be generated and sent to the staff terminal.

[0101] There are various ways to send alerts, aiming to ensure that alert information can be conveyed to relevant personnel in a timely and accurate manner. The following are some common ways to send alerts:

[0102] Email Alert: Alert information is sent to relevant personnel via email. This method is suitable for alert information that needs to be recorded or reviewed later.

[0103] SMS alert: Alert information is sent to a designated mobile phone number via SMS. This method is suitable for emergency alerts that require immediate action or response.

[0104] WeChat Alarm: Send alarm information to WeChat and notify through WeChat official account, enterprise WeChat or personal WeChat.

[0105] Telephone alarm: Directly notify relevant personnel via telephone voice. This method is usually used for the most urgent alarm situations.

[0106] API push: Provides an external API, allowing third-party systems to periodically call the API to pull alarm events. This method is suitable for scenarios where alarm information needs to be integrated into existing systems.

[0107] Message queue push: Alarm information is placed in a message queue, and consumers retrieve the alarm from the queue and then notify the user. This method helps decouple services and asynchronously process alarm events.

[0108] WebSocket / SSE push: Maintain a persistent connection via the WebSocket or SSE (Server-Sent Events) protocol to push real-time messages from the server to the client.

[0109] When choosing an alert delivery method, consider the urgency of the alert, the characteristics of the recipient, and the actual needs of the system. For example, for urgent alerts requiring immediate action, SMS, phone calls, or WebSocket / SSE push methods can be used; for alerts that require awareness but don't require immediate action, email or WeChat can be used; and for scenarios where alert information needs to be integrated into existing systems, API push or message queue push can be used.

[0110] This embodiment can realize batch extraction of table-based compensation scopes in agricultural insurance clauses, and intelligently check consistency and issue alarms.

[0111] After step S500, the following steps are also included:

[0112] generating an agricultural insurance clause comparison report based on the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards;

[0113] Based on the agricultural insurance clause comparison report, the duplication of compensation standards is marked.

[0114] During specific implementation, the duplication between the target compensation standard and several original agricultural insurance compensation standards is obtained, and a comparison report is generated based on the duplication. In the comparison report, the repeated or similar clauses in the agricultural insurance clause document are marked.

[0115] In this embodiment, the automated processing of table extraction and comparison reduces the clause review time by more than 80%, shortening the average review time for each clause from 11 hours to less than 20 minutes.

[0116] After marking the duplication of compensation standards based on the agricultural insurance clause comparison report, it also includes:

[0117] Generate a proposed clause revision plan based on the duplication between the target compensation standard and the several original agricultural insurance compensation standards;

[0118] The duplication between the target compensation standard and several original agricultural insurance compensation standards is displayed.

[0119] Based on the data verification results, a proposed clause revision plan is automatically generated to reduce the duplication of compensation scope and compensation ratio. Visual tools can also be provided to help business personnel intuitively view and analyze the comparison results. Furthermore, the data verification process can be integrated into the agricultural insurance product clause declaration business process to improve the efficiency of clause management and reduce labor costs. In some other embodiments, it can also be used in conjunction with the agricultural insurance core system and product factory to intelligently identify and intercept in advance. Using its semantic understanding ability, it can perform semantic verification on the compensation scope of different products of different institutions, identify the repeated content of the compensation scope and issue an alarm. If there is repeated content, it will prompt for modification, and locate and prompt similar clauses. Reduce conflicts due to repeated clauses.

[0120] Compared with the existing technology, the data verification method of the embodiment of the present invention has the following technical advantages: the automated processing of the extraction and comparison of the compensation standards of agricultural insurance clause documents reduces the clause review time by more than 80%, from the original average of 11 hours for each clause review to less than 20 minutes. The automated semantic similarity calculation reduces the workload of manual review and saves labor costs; when there is duplication in the agricultural insurance compensation standards, an alarm message is generated, and the staff can modify the agricultural insurance clauses based on the alarm message, thereby reducing the duplication and conflict of clauses and saving operating costs.

[0121] It should be noted that there is not necessarily a certain order between the above steps. A person skilled in the art can understand, based on the description of the embodiments of the present invention, that in different embodiments, the above steps may have different execution orders, that is, they may be executed in parallel, or may be executed interchangeably, etc.

[0122] Another embodiment of the present invention provides a data verification device, which corresponds one-to-one to the data verification method of the above embodiment. Figure 7 As shown, the device 1 includes a data acquisition module 100, a data conversion module 200, a data storage module 300, a data verification module 400 and an alarm information generation module 500. The functional modules are described in detail as follows:

[0123] The data acquisition module 100 is used to detect the agricultural insurance clause document filing instruction and obtain the agricultural insurance clause document to be filed;

[0124] The data conversion module 200 is used to identify the compensation standard of the agricultural insurance clause document to be reported, and convert the identified compensation standard into structured data, which is recorded as the target compensation standard;

[0125] The data storage module 300 is used to store the target compensation standard in a preset agricultural insurance compensation standard database, wherein the preset agricultural insurance compensation standard database stores a plurality of original agricultural insurance compensation standards;

[0126] A data verification module 400 is used to calculate the semantic similarity between the target compensation standard and the plurality of original agricultural insurance compensation standards, and verify the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards based on the semantic similarity;

[0127] The warning information generating module 500 is used to generate a warning message if the target compensation standard is repeated with the plurality of original agricultural insurance compensation standards.

[0128] The specific implementation method is shown in the method embodiment, which will not be repeated here.

[0129] In one embodiment, the data conversion module 200 is specifically configured to:

[0130] Identifying the compensation standard of the agricultural insurance clause document to be reported based on the multimodal large model to obtain the identified compensation standard;

[0131] The identified compensation standard is converted into structured data in a preset format and recorded as the target compensation standard.

[0132] The specific implementation method is shown in the method embodiment, which will not be repeated here.

[0133] In one embodiment, the data storage module 300 is specifically configured to:

[0134] Based on the agricultural insurance clause document to be filed, obtaining the issuing institution corresponding to the target compensation standard;

[0135] The target compensation standard is stored in a preset agricultural insurance compensation standard database according to the issuing institution, wherein the preset agricultural insurance compensation standard database stores a plurality of original agricultural insurance compensation standards.

[0136] The specific implementation method is shown in the method embodiment, which will not be repeated here.

[0137] In one embodiment, the data verification module 400 is specifically configured to:

[0138] Inputting the target compensation standard and the several original agricultural insurance compensation standards into a semantic similarity calculation model;

[0139] The semantic similarity output by the semantic similarity calculation model is obtained, and the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards is verified according to the semantic similarity.

[0140] The specific implementation method is shown in the method embodiment, which will not be repeated here.

[0141] In one embodiment, the data verification module 400 is further configured to:

[0142] Obtaining semantic similarity output by the semantic similarity calculation model, wherein the semantic similarity calculation model is a deep learning model;

[0143] Determine whether the semantic similarity is greater than or equal to a preset similarity threshold;

[0144] If the semantic similarity is greater than or equal to a preset similarity threshold, it is determined that the target compensation standard and the plurality of original agricultural insurance compensation standards are duplicated;

[0145] If the semantic similarity is less than a preset similarity threshold, it is determined that there is no duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards.

[0146] The specific implementation method is shown in the method embodiment, which will not be repeated here.

[0147] In one embodiment, the apparatus further comprises a marking module, the marking module being configured to:

[0148] generating an agricultural insurance clause comparison report based on the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards;

[0149] Based on the agricultural insurance clause comparison report, the duplication of compensation standards is marked.

[0150] The specific implementation method is shown in the method embodiment, which will not be repeated here.

[0151] In one embodiment, the device further comprises a display module, the display module being configured to:

[0152] Generate a proposed clause revision plan based on the duplication between the target compensation standard and the several original agricultural insurance compensation standards;

[0153] The duplication between the target compensation standard and several original agricultural insurance compensation standards is displayed.

[0154] The specific implementation method is shown in the method embodiment, which will not be repeated here.

[0155] The present invention provides a data verification device that automatically extracts and compares the compensation standard tables in agricultural insurance clauses, automatically compares and verifies, issues intelligent alarms for homogeneous compensation ranges, and automatically calculates semantic similarity, thereby reducing the workload of manual review and saving labor costs. When there are duplications in agricultural insurance compensation standards, an alarm message is generated, and staff can modify the agricultural insurance clauses based on the alarm message, thereby reducing duplications and conflicts in clauses and saving operating costs.

[0156] Another embodiment of the present invention provides a computer device, which may be a server, and its internal structure diagram may be as shown in FIG. Figure 8As shown. The computer device includes a processor, memory, network interface and database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external client via a network connection. When the computer program is executed by the processor, it implements the functions or steps on the server side of a data verification method.

[0157] In one embodiment, a computer device is provided. The computer device may be a client, and its internal structure diagram may be as follows: Figure 9 As shown. The computer device includes a processor, memory, network interface, display screen, and input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server via a network connection. When executed by the processor, the computer program implements the functions or steps on the client side of a data verification method.

[0158] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the following steps are performed:

[0159] Detects the agricultural insurance clause document filing instruction and obtains the agricultural insurance clause document to be filed;

[0160] Identifying the compensation standard of the agricultural insurance clause document to be reported, and converting the identified compensation standard into structured data, which is recorded as the target compensation standard;

[0161] Storing the target compensation standard in a preset agricultural insurance compensation standard database, wherein the preset agricultural insurance compensation standard database stores a plurality of original agricultural insurance compensation standards;

[0162] Calculating semantic similarities between the target compensation standard and the plurality of original agricultural insurance compensation standards respectively, and verifying duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards based on the semantic similarities;

[0163] If the target compensation standard is repeated with the several original agricultural insurance compensation standards, an alarm message is generated.

[0164] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0165] Detects the agricultural insurance clause document filing instruction and obtains the agricultural insurance clause document to be filed;

[0166] Identifying the compensation standard of the agricultural insurance clause document to be reported, and converting the identified compensation standard into structured data, which is recorded as the target compensation standard;

[0167] Storing the target compensation standard in a preset agricultural insurance compensation standard database, wherein the preset agricultural insurance compensation standard database stores a plurality of original agricultural insurance compensation standards;

[0168] Calculating semantic similarities between the target compensation standard and the plurality of original agricultural insurance compensation standards respectively, and verifying duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards based on the semantic similarities;

[0169] If the target compensation standard is repeated with the several original agricultural insurance compensation standards, an alarm message is generated.

[0170] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or computer device can be found in the relevant descriptions of the server side and the client side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0171] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchl ink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0172] The embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules may be selected to achieve the objectives of the present embodiments as needed.

[0173] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a general hardware platform, or of course, can also be implemented by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the relevant technology can be embodied in the form of a software product. This computer software product can be present in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or certain parts of the embodiment.

[0174] It should be noted that if software tools or components other than those of our company appear in the embodiments of this application, they are only used for illustration and do not represent actual use.

[0175] Conditional language such as "can," "may," or "might," among others, unless specifically stated otherwise or otherwise understood within the context as used, is generally intended to convey that particular embodiments can include, while other embodiments do not, particular features, elements, and / or operations. Thus, such conditional language is also generally intended to imply that features, elements, and / or operations are anyway required for one or more embodiments or that one or more embodiments must include logic for determining, with or without input or prompting, whether such features, elements, and / or operations are included or to be performed in any particular embodiment.

[0176] What has been described herein in this specification and the accompanying drawings includes examples that can provide data verification methods and devices. Of course, it is not possible to describe every conceivable combination of elements and / or methods for the purpose of describing the various features of the present disclosure, but it is recognized that many other combinations and permutations of the disclosed features are possible. Therefore, it is apparent that various modifications can be made to the present disclosure without departing from the scope or spirit of the present disclosure. In addition, or in an alternative, other embodiments of the present disclosure may be apparent from consideration of this specification and the accompanying drawings and from the practice of the present disclosure as presented herein. It is intended that the examples proposed in this specification and the accompanying drawings are considered to be illustrative and not restrictive in all respects. Although specific terms are employed herein, they are used in a general and descriptive sense and are not used for the purpose of limitation.

Claims

1. A data verification method, characterized in that , the method comprises: Detects the agricultural insurance clause document filing instruction and obtains the agricultural insurance clause document to be filed; Identifying the compensation standard of the agricultural insurance clause document to be reported, and converting the identified compensation standard into structured data, which is recorded as the target compensation standard; Storing the target compensation standard in a preset agricultural insurance compensation standard database, wherein the preset agricultural insurance compensation standard database stores a plurality of original agricultural insurance compensation standards; Calculating semantic similarities between the target compensation standard and the plurality of original agricultural insurance compensation standards respectively, and verifying duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards based on the semantic similarities; If the target compensation standard and the several original agricultural insurance compensation standards are repeated, an alarm message is generated; The process of identifying the compensation standard of the agricultural insurance clause document to be filed and converting the identified compensation standard into structured data, which is recorded as the target compensation standard, includes: Identifying the compensation standard of the agricultural insurance clause document to be reported based on the multimodal large model to obtain the identified compensation standard; Converting the identified compensation standard into structured data in a preset format, which is recorded as the target compensation standard; The target compensation standard is stored in a preset agricultural insurance compensation standard database, wherein the preset agricultural insurance compensation standard database stores a plurality of original agricultural insurance compensation standards: Based on the agricultural insurance clause document to be filed, obtaining the issuing institution corresponding to the target compensation standard; Storing the target compensation standard in a preset agricultural insurance compensation standard database according to the issuing institution, wherein the preset agricultural insurance compensation standard database stores a plurality of original agricultural insurance compensation standards; The respectively calculating the semantic similarity between the target compensation standard and the plurality of original agricultural insurance compensation standards, and verifying the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards according to the semantic similarity, includes: Inputting the target compensation standard and the several original agricultural insurance compensation standards into a semantic similarity calculation model; Obtaining the semantic similarity output by the semantic similarity calculation model, and verifying the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards based on the semantic similarity; The semantic similarity calculation model is a deep learning model.

2. The data verification method according to claim 1, wherein: The obtaining of the semantic similarity output by the semantic similarity calculation model and verifying the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards according to the semantic similarity include: Obtaining the semantic similarity output by the semantic similarity calculation model; Determine whether the semantic similarity is greater than or equal to a preset similarity threshold; If the semantic similarity is greater than or equal to a preset similarity threshold, it is determined that the target compensation standard and the plurality of original agricultural insurance compensation standards are duplicated; If the semantic similarity is less than a preset similarity threshold, it is determined that there is no duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards.

3. The data verification method according to any one of claims 1 to 2, characterized in that: If the target compensation standard is repeated with the several original agricultural insurance compensation standards, after generating the alarm information, the method further includes: generating an agricultural insurance clause comparison report based on the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards; Based on the agricultural insurance clause comparison report, the duplication of compensation standards is marked.

4. The data verification method according to claim 3, wherein: After marking the duplication of compensation standards based on the agricultural insurance clause comparison report, it also includes: Generate a proposed clause revision plan based on the duplication between the target compensation standard and the several original agricultural insurance compensation standards; The duplication between the target compensation standard and several original agricultural insurance compensation standards is displayed.

5. A data verification device, characterized in that: The device comprises: A data acquisition module is used to detect an instruction to report agricultural insurance clause documents and obtain the agricultural insurance clause documents to be reported; A data conversion module is used to identify the compensation standard of the agricultural insurance clause document to be reported, and convert the identified compensation standard into structured data, which is recorded as the target compensation standard; A data storage module, configured to store the target compensation standard in a preset agricultural insurance compensation standard database, wherein the preset agricultural insurance compensation standard database stores a plurality of original agricultural insurance compensation standards; a data verification module, configured to respectively calculate the semantic similarity between the target compensation standard and the plurality of original agricultural insurance compensation standards, and verify the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards based on the semantic similarity; an alarm information generating module, configured to generate an alarm message if the target compensation standard and the plurality of original agricultural insurance compensation standards are duplicated; The process of identifying the compensation standard of the agricultural insurance clause document to be filed and converting the identified compensation standard into structured data, which is recorded as the target compensation standard, includes: Identifying the compensation standard of the agricultural insurance clause document to be reported based on the multimodal large model to obtain the identified compensation standard; Converting the identified compensation standard into structured data in a preset format, which is recorded as the target compensation standard; The target compensation standard is stored in a preset agricultural insurance compensation standard database, wherein the preset agricultural insurance compensation standard database stores a plurality of original agricultural insurance compensation standards, including: Based on the agricultural insurance clause document to be filed, obtaining the issuing institution corresponding to the target compensation standard; Storing the target compensation standard in a preset agricultural insurance compensation standard database according to the issuing institution, wherein the preset agricultural insurance compensation standard database stores a plurality of original agricultural insurance compensation standards; The respectively calculating the semantic similarity between the target compensation standard and the plurality of original agricultural insurance compensation standards, and verifying the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards according to the semantic similarity, includes: Inputting the target compensation standard and the several original agricultural insurance compensation standards into a semantic similarity calculation model; Obtaining the semantic similarity output by the semantic similarity calculation model, and verifying the duplication between the target compensation standard and the plurality of original agricultural insurance compensation standards based on the semantic similarity; The semantic similarity calculation model is a deep learning model.

6. A computer device, characterized in that: The computer device includes at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can perform the steps of the data verification method according to any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions. When the computer-executable instructions are executed by one or more processors, the one or more processors can execute the steps of the data verification method according to any one of claims 1 to 4.

Images