A remote notarization document generation method for automatic matching business

Abstract

The present invention discloses a remote notarial document generation method for automatic matching services, comprising: obtaining a notarial request submitted by an applicant, obtaining a notarial matter type according to the notarial request analysis, matching a corresponding notarial document template according to the notarial matter type, and obtaining an information tag according to the notarial document template; obtaining applicant information by obtaining the information tag, generating a personalized question group according to the applicant information and the notarial matter type, conducting remote interactive question and answer according to the personalized question group, and collecting the applicant's question and answer video; converting the voice in the applicant's question and answer video into question and answer text, grabbing the question and answer text content according to the information tag and filling it into the notarial document template to generate a notarial document; converting the notarial document into encrypted information, embedding the encrypted information into each frame of the question and answer video in sequence to obtain an encrypted notarial video, and uploading the encrypted notarial video to a server for storage.

Description

Technical Field

[0001] The present invention relates to the technical field of notarial document generation, and in particular to a remote notarial document generation method for automatic matching services. Background Art

[0002] Remote notarization is a notarization service performed by electronic means. Instead of going to the notary office in person, the parties can communicate and verify with the notary remotely through video conferencing software or other online platforms. Remote notarization usually involves the following steps: identity verification through video conferencing and identity documents to confirm the identity of the parties; the notary reviews the documents to be notarized online to ensure their integrity and legality; the notary conducts a real-time video conference with the parties to verify information, answer questions, and complete other necessary parts of the notarization process; the parties complete the signing of the document through electronic signature to ensure the legal effect of the document; the electronic version of the notarial certificate is automatically generated based on the content of the conversation and provided to the parties.

[0003] On the one hand, during the entire remote notarization process, the notary needs to participate in the question-and-answer process and generate the notarization text according to the applicant's needs. This remote notarization method is cumbersome and requires a lot of manual participation. In addition, the fully automated notarization problem is not suitable for the actual situation of the applicant; on the other hand, during the entire remote notarization process, the corresponding notarization output files involve multiple file types such as videos, images, and texts. The notarization output files usually need to upload the notarization videos and materials to the notary office or the upper cloud storage. These output files, as evidence of the legality of the notarization, have high requirements for the confidentiality and integrity of the transmission and storage links. At present, in the evidence transmission and storage link in the field of remote notarization, the main method is to upload and store each file separately. Although packaging and secure encryption technologies are used for storage and transmission to varying degrees, file separation causes the attack points to be dispersed, and the attack surface of man-in-the-middle attacks and rainbow table decryption of single files is expanded.

[0004] The purpose of the present invention is to design a remote notarization document generation method for automatic matching services in view of the above-mentioned problems in the prior art. Summary of the invention

[0005] In view of this, the purpose of the present invention is to propose a remote notarization document generation method for automatic matching services, which can solve the above-mentioned problems.

[0006] The present invention provides a remote notarization document generation method for automatic matching services, comprising:

[0007] Obtain the notarization request submitted by the applicant, parse the notarization request to obtain the notarization item type, match the corresponding notarization document template according to the notarization item type, and obtain the information tag according to the notarization document template;

[0008] By obtaining the applicant information obtained through the information tag, a personalized question group is generated according to the applicant information and the type of notarization matter, and a remote interactive question and answer is conducted according to the personalized question group, and the applicant's question and answer video is collected;

[0009] Convert the voice in the applicant's Q&A video into Q&A text, grab the Q&A text content according to the information tag and fill it into the notarization document template to generate a notarization document;

[0010] The notarization document is converted into encrypted information, and the encrypted information is sequentially embedded into each frame of the question-and-answer video to obtain an encrypted notarization video, which is then uploaded to the server for storage.

[0011] Furthermore, the notarization request submitted by the applicant is obtained, and the notarization item type obtained by parsing the notarization request includes:

[0012] Performing text preprocessing on the notarization request to obtain a preprocessed notarization request text;

[0013] Extract keywords and their associations from the notarization request text;

[0014] The keywords of the notarization request text and their associations are input into a pre-trained notarization item type classification model to obtain the notarization item type.

[0015] Furthermore, the keywords and their associations extracted from the notarization request text include:

[0016] Extracting the keywords of the matter in the notarization request text by using TF-IDF technology;

[0017] Extracting entity type keywords in the notarization request text by using a named entity recognition model;

[0018] Analyze the dependency relationship between matter keywords and entity type keywords to obtain keyword association relationships.

[0019] Furthermore, matching a corresponding notarial document template according to the notarial matter type and obtaining an information tag according to the notarial document template includes:

[0020] According to the type of notarization matter, matching a suitable notarization document template from a predefined template library;

[0021] An information tag is extracted from the notarial document template by using a field identifier.

[0022] Furthermore, the applicant information obtained by obtaining the information tag, generating a personalized question group according to the applicant information and the notarization item type includes:

[0023] Using the information tag as the information index to be obtained, searching the notarization request submitted by the applicant through the information index to obtain the applicant information;

[0024] Establish a standard question and answer template according to the notarial matter type, and align it with the information tag to form a standard question group;

[0025] The standard question set is dynamically adjusted according to the applicant information to obtain a personalized question set.

[0026] Furthermore, the applicant information includes: gender, age, cultural and educational background, place of residence, and location of the notarization matter.

[0027] Furthermore, the method of dynamically adjusting the standard question set according to the applicant information to obtain the personalized question set includes:

[0028] Input applicant information and question set into a pre-trained prediction model to obtain a predicted answer for each question in the question set;

[0029] The question branch structure in the standard question group is dynamically adjusted according to the predicted answer of each question to obtain a personalized question group.

[0030] Furthermore, the pre-trained prediction model is obtained by the following steps:

[0031] Gender, age, cultural and educational background, place of residence, location of notarization items, and issues concerning notarization items of several applicants are collected, and the gender, age, cultural and educational background, place of residence, location of notarization items, and issues concerning notarization items of the applicants are divided into several samples, and the answers of each sample are calibrated to form a prediction sample training set;

[0032] The prediction model is constructed by a neural network model, and the prediction model is trained by a prediction sample training set to obtain the parameters of the prediction model.

[0033] Furthermore, the notarization document is converted into encrypted information, and the encrypted information is sequentially embedded into each frame of the question-and-answer video to obtain the encrypted notarization video, including:

[0034] Convert the notarized document into a byte stream, encrypt the byte stream using an encryption algorithm to generate encrypted information, and convert the encrypted information into a binary format;

[0035] Decoding the question-and-answer video to obtain a video frame sequence of the question-and-answer video, and encrypting the video frame sequence using a key to generate a pseudo-random sequence;

[0036] The binary bits of the encrypted information are embedded into the least significant bits of the pixels of each video frame in a pseudo-random sequence, and the embedded video frames are re-encoded to generate the encrypted notarized video.

[0037] Further, embedding the binary bits of the encryption information into the least significant bits of the pixels of each video frame in a sequential order comprises:

[0038] Traverse each video frame and select the least significant bit of any RGB channel in each video frame;

[0039] Replace the least significant bit with a binary bit of the encrypted message.

[0040] Beneficial effects of the present invention:

[0041] First, the notarization item type is automatically identified through the applicant's notarization request, and the corresponding notarization document template is matched. The automated template matching reduces the error of manual selection, improves the efficiency and accuracy of document generation. At the same time, information tags are obtained through the notarization document template for subsequent filling in of text.

[0042] Secondly, a group of questions for remote video interaction is automatically generated based on the notarization matters. This solves the problem of time and resources consumed in manually planning complex question and answer content. By combining automatic templates and real-time questions and answers, personalized services are provided while ensuring the uniformity and consistency of the notarization procedures.

[0043] The third is to fill in the notarization document through the voice content in the question-and-answer video combined with the previous information tags to improve the speed and completeness of the document and ensure that the application information is accurately reflected in the notarization document.

[0044] Fourth, by embedding the encrypted information generated by the notarized document text in the video frame, the security risks brought by the scattered transmission of different files are reduced. The notarized document is combined with the video to reduce the attack points in the transmission and enhance the ability to protect against man-in-the-middle attacks and other common attack methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description 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 creative work.

[0046] Figure 1 It is a flow chart of the method of embodiment 1. DETAILED DESCRIPTION

[0047] To facilitate understanding by those skilled in the art, the structure of the present invention is now further described in detail with reference to the embodiments and the accompanying drawings. It should be understood that the steps mentioned in the present embodiment, unless otherwise specified in their order, can be adjusted in order according to actual needs, and can even be executed simultaneously or partially simultaneously.

[0048] like Figure 1 As shown, an embodiment of the present invention provides a method for generating a remote notarization document for an automatic matching service, comprising:

[0049] S1 obtains the notarization request submitted by the applicant, parses the notarization request to obtain the notarization item type, matches the corresponding notarization document template according to the notarization item type, and obtains the information tag according to the notarization document template;

[0050] S101 performs text preprocessing on the notarization request to obtain a preprocessed notarization request text;

[0051] S102 extracts keywords and their associations from the notarization request text;

[0052] S1021 extracts the keywords of the matter in the notarization request text by using TF-IDF technology;

[0053] S1022 extracts entity type keywords in the notarization request text through a named entity recognition model;

[0054] S1023 analyzes the dependency relationship between the matter keywords and the entity type keywords to obtain the keyword association relationship.

[0055] S103 inputs the keywords of the notarization request text and their associations into a pre-trained notarization item type classification model to obtain the notarization item type.

[0056] In this step, the text content provided by the applicant generally includes: basic information of the applicant: name, contact information, address, etc.; description of notarization matters: specific matters that need to be notarized, such as real estate sales, power of attorney, document certification, etc.; background information: background information about the request, which may include the specific circumstances of the incident or judicial procedure; requirements and instructions: the applicant's specific needs or expectations for the notarization results.

[0057] Preprocessing the text includes removing invalid characters, punctuation, stop words, and unifying the text format, which can effectively reduce noise and standardize text data for subsequent analysis. The TF-IDF (Term Frequency-Inverse Document Frequency) method is used to identify words that frequently appear in the notarization request text but are not common in other texts. These words usually have high relevance and can reveal the text theme. Use the named entity recognition (NER) model to identify specific types of entities (such as names, organizations, and places) from the text. These entities help understand the specific context and subject of the request. By parsing the grammatical relationship and dependency structure between keywords, an association graph between keywords is constructed to help understand the logical organization and semantic connection between keywords. The extracted keywords and their associations are input into the pre-trained classification model to automatically determine the corresponding notarization item type. The classification model can be a model based on machine learning (such as SVM, random forest) or deep learning (such as BERT).

[0058] S104, matching a suitable notarial document template from a predefined template library according to the notarial matter type;

[0059] S105 extracts information tags from the notarization document template through field identifiers.

[0060] In this step, once the type of notarization application is determined (through the previous analysis and classification steps), the document template that matches this type can be found in the template library. The predefined template library contains standard document formats for different notarization matters. For example, common notarization matters such as sales contracts, power of attorney, and wills each have their own corresponding templates. The corresponding template is automatically selected according to the type of notarization matter. Through the use of the template library, it can be ensured that the format and content of the generated notarization document meet the predefined standards. The template usually contains specific field identifiers, which are used to indicate the key information that needs to be filled in or generated, such as the applicant's name, date, specific terms, etc. These field identifiers (information tags) are extracted from the selected template. These tags will help automatically fill in the corresponding parts of the notarization document. Through the above steps, the applicant's notarization needs and the corresponding notarization type template and information tags can be quickly and accurately obtained.

[0061] S2 obtains applicant information through information tags, generates a personalized question group based on the applicant information and notarization matter type, conducts remote interactive Q&A based on the personalized question group, and collects applicant Q&A videos;

[0062] S201 uses the information tag as the information index to be obtained, retrieves the notarization request submitted by the applicant through the information index, and obtains the applicant information;

[0063] Furthermore, the applicant information includes: gender, age, cultural and educational background, place of residence, and location of the notarization matter.

[0064] In this step, the information tag is used as an index identifier to retrieve the notarization request application text and extract the applicant's specific information, such as gender, age, cultural and educational background, place of residence, location of the notarization matter, etc., for subsequent prediction of the applicant's answer.

[0065] S202: establishing a standard question and answer template according to the notarial matter type, and aligning it with the information tag to form a standard question group;

[0066] In this step, notarization matters can be divided into the following main types: Signature notarization: including notarization of signatures of documents such as authorizations, declarations, deeds, contracts and agreements. This notarization ensures the authenticity of the signatures on the documents and the true intentions of the parties. Identity notarization: used to prove the authenticity of identity information, such as identity certificates, proof of kinship, passport identity authentication, etc. Property notarization: involving the ownership and transfer of property, such as real estate sales, wills, inheritance rights, gift contracts, etc. Behavior notarization: involving the proof of specific behaviors, such as unilateral statements, promises to perform, proof of attendance or the authenticity of showing a certain behavior. Document notarization: verification of original documents and authenticity certification of document copies, such as notarization of copies of academic certificates, business licenses, company articles of association, and contract texts. Marriage and family notarization: involving marital status, property agreements, husband and wife agreements, child support, maintenance agreements, etc. Company business notarization: covering all kinds of notarization matters related to company business, such as company registration, use of official seals, shareholders' meeting resolutions, commercial contracts, public trust and other related documents. Entrustment notarization: notarization of authorizing others to perform legal acts on their behalf, such as entrusting the provision of legal services, buying and selling real estate, handling banking affairs, etc. Declaration and certification notarization: used for personal declarations of certain facts, such as a certificate of no criminal record, a certificate of unmarried status, a certificate of income, etc. Translation notarization: ensuring the accuracy of translated documents, which can be used for contract translation, official document translation, and certification document translation. Maritime notarization: notarization of legal matters related to the ocean or shipping, such as ship mortgages, notarization reliance on maritime disputes, and supply contracts. Intellectual property notarization: notarization of applications, transfers, and other legal matters regarding patents, trademarks, copyrights, etc. Different types of notarization matters correspond to different notarization templates, so the corresponding notarization questions and answers are also different. Create standardized question and answer templates for each type of notarization matter, and align the templates with information tags to ensure that all relevant information is correctly extracted.

[0067] S203 dynamically adjusts the standard question set according to the applicant information to obtain a personalized question set.

[0068] S2031 inputs the applicant information and the question group into a pre-trained prediction model to obtain a predicted answer for each question in the question group;

[0069] Furthermore, the pre-trained prediction model is obtained through the following steps:

[0070] Gender, age, cultural and educational background, place of residence, location of notarization items, and issues concerning notarization items of several applicants are collected, and the gender, age, cultural and educational background, place of residence, location of notarization items, and issues concerning notarization items of the applicants are divided into several samples, and the answers of each sample are calibrated to form a prediction sample training set;

[0071] The prediction model is constructed by a neural network model, and the prediction model is trained by a prediction sample training set to obtain the parameters of the prediction model.

[0072] S2032 dynamically adjusts the question branch structure in the standard question group according to the predicted answer to each question to obtain a personalized question group.

[0073] In this step, a pre-trained neural network model is used to predict the potential answer to each question based on the applicant information and the standard question set. The question branch structure in the question set is modified based on the predicted answer to ensure that the order and content of the questions match the applicant's background and expected response, thereby forming a personalized question set. By automatically responding to individual needs, the relevance of the question-and-answer process is improved, simplifying and accelerating the notarization process. For example, an applicant is undergoing a remote video notarization procedure related to a real estate transfer. During this notarization process, the notary needs to ensure that the applicant understands and agrees to all aspects of the transaction. If the applicant has a low level of education, it is necessary to add a lot of detailed explanation questions and add answers to explain the relevant laws and regulations for the applicant's actual question-and-answer time.

[0074] S3 converts the voice in the applicant's Q&A video into Q&A text, grabs the Q&A text content according to the information tag, fills it into the notarization document template, and generates a notarization document;

[0075] In this step, based on the personalized question set obtained in the previous step, a virtual notary and the applicant can have a video Q&A conversation, and voice recognition technology can be used to convert the voice of the conversation between the applicant and the virtual notary in the video Q&A into text. After the voice is converted to text, the text needs to be organized, including punctuation correction, format normalization, etc., to ensure the coherence and accuracy of the text. Use the information tags in the notarial document template to identify and extract key information in the text, such as name, date, commission content and other specific information. Automatically fill in the corresponding notarial document template based on the content extracted from the Q&A text. Templated documents can ensure the consistency of file structure and content.

[0076] S4 converts the notarization document into encrypted information, embeds the encrypted information into each frame of the question-and-answer video in sequence, obtains the encrypted notarization video, and uploads the encrypted notarization video to the server for storage.

[0077] S401 converts the notarized document into a byte stream, encrypts the byte stream using an encryption algorithm to generate encrypted information, and converts the encrypted information into a binary format;

[0078] In this step, the notarized document is converted into a byte stream, and then the encryption algorithm is applied to encrypt it, generating encrypted information, and converting it into a binary format. This ensures that the notarized document is encrypted during transmission and storage to protect confidentiality. The encryption algorithm can use the hash algorithm SHA-256 or MD5.

[0079] S402 decodes the question-and-answer video to obtain a video frame sequence of the question-and-answer video, and encrypts the video frame sequence using a key to generate a pseudo-random sequence;

[0080] In this step, the question-and-answer video is decoded to obtain each frame of the video. At the same time, the video frame is encrypted with a key to generate a pseudo-random sequence. This process reduces the predictability of the order of the video frames, thereby enhancing security, rather than simply using the sequential frame sequence of the video.

[0081] S403 embeds the binary bits of the encryption information into the least significant bits of the pixels of each video frame in a pseudo-random sequence, and re-encodes the embedded video frames to generate encrypted notarized videos.

[0082] S4031 traverses each video frame and selects the least significant bit of any RGB channel in each video frame;

[0083] S4032 replaces the least significant bit with a binary bit of encrypted information.

[0084] In this step, the binary bits of the encrypted information are embedded into the video frames by replacing the least significant bit (Least Significant Bit) of the RGB channel in each frame. The least significant bit is chosen because changing this bit has the least impact on the visual effect of the image or video, but can carry information.

[0085] Transmitting the encrypted question-and-answer video of the notarized document to the server can effectively avoid uploading and storing each file separately, solving the problem of scattered attack points caused by file separation. The encrypted notarized video obtained on the server can be decoded, and the same pseudo-random sequence can be used to reversely calculate the video frame order for information extraction, extract information bits from specified frames and pixel positions, and synthesize a complete binary stream, which can then be decrypted back to the original information.

[0086] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0087] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0088] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0089] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0090] It should be noted that in the claims, any reference signs placed between brackets shall not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claim. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention may be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third etc. does not indicate any order. These words may be interpreted as names.

[0091] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0092] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

[0093] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0094] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms should not be understood as necessarily referring to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

Claims

1. A remote notarization document generation method for automatic matching services, characterized in that: include: Obtain the notarization request submitted by the applicant, parse the notarization request to obtain the notarization item type, match the corresponding notarization document template according to the notarization item type, and obtain the information tag according to the notarization document template; The applicant information is obtained by obtaining information tags, and a personalized question group is generated according to the applicant information and the type of notarization matter. A remote interactive question and answer is conducted according to the personalized question group, and the applicant's question and answer video is collected. Specifically: The information tag is used as the information index to be obtained, and the notarization request submitted by the applicant is retrieved through the information index to obtain the applicant information, wherein the applicant information includes: gender, age, cultural and educational background, place of residence, and location of the notarization matter; Establish a standard question and answer template according to the notarial matter type, and align it with the information tag to form a standard question group; Dynamically adjust the standard question set according to the applicant information to obtain a personalized question set. Specifically, the applicant information and the question set are input into a pre-trained prediction model to obtain a predicted answer for each question in the question set. Dynamically adjust the question branch structure in the standard question group according to the predicted answer of each question to obtain a personalized question group; Convert the voice in the applicant's Q&A video into Q&A text, grab the Q&A text content according to the information tag and fill it into the notarization document template to generate a notarization document; Convert the notarization document into encrypted information, embed the encrypted information into each frame of the question-and-answer video in sequence, obtain the encrypted notarization video, and upload the encrypted notarization video to the server for storage. Specifically: Convert the notarized document into a byte stream, encrypt the byte stream using an encryption algorithm to generate encrypted information, and convert the encrypted information into a binary format; Decoding the question-and-answer video to obtain a video frame sequence of the question-and-answer video, and encrypting the video frame sequence using a key to generate a pseudo-random sequence; The binary bits of the encrypted information are embedded into the least significant bits of the pixels of each video frame in a pseudo-random sequence, and the embedded video frames are re-encoded to generate the encrypted notarized video. Specifically: Traverse each video frame and select the least significant bit of any RGB channel in each video frame; Replace the least significant bit with a binary bit of the encrypted message.

2. According to claim 1, a remote notarization document generation method for automatic matching services is characterized in that: The notarization request submitted by the applicant is obtained, and the notarization item types obtained by parsing the notarization request include: Performing text preprocessing on the notarization request to obtain a preprocessed notarization request text; Extract keywords and their associations from the notarization request text; The keywords of the notarization request text and their associations are input into a pre-trained notarization item type classification model to obtain the notarization item type.

3. The method for generating remote notarization documents for automatic matching services according to claim 2, characterized in that: The keywords and their associations for extracting the notarization request text include: Extracting keywords from the notarization request text by using TF-IDF technology; Extracting entity type keywords in the notarization request text by using a named entity recognition model; Analyze the dependency relationship between matter keywords and entity type keywords to obtain keyword association relationships.

4. The method for generating remote notarization documents for automatic matching services according to claim 1, characterized in that: The matching of the notarial document template to the notarial item type and obtaining the information tag according to the notarial document template includes: According to the type of notarization matter, matching a suitable notarization document template from a predefined template library; An information tag is extracted from the notarial document template by using a field identifier.

5. The method for generating remote notarization documents for automatic matching services according to claim 1, characterized in that: The pre-trained prediction model is obtained by the following steps: Gender, age, cultural and educational background, place of residence, location of notarization items, and issues concerning notarization items of several applicants are collected, and the gender, age, cultural and educational background, place of residence, location of notarization items, and issues concerning notarization items of the applicants are divided into several samples, and the answers of each sample are calibrated to form a prediction sample training set; The prediction model is constructed by a neural network model, and the prediction model is trained by a prediction sample training set to obtain the parameters of the prediction model.

Images