Knowledge intelligent classification label generation method and system for cloud document

Abstract

The invention discloses a knowledge intelligent classification tag generation method and system for a cloud document, and the method comprises the steps: carrying out the document information transmission in a secure transmission encryption mode, transmitting a cloud to a local document, and building a cloud local document database; a data synchronization and coding mode is adopted, a cloud local document directory is synchronized with a file, and a cloud local document database is updated; the text information is processed for information classification, classification labels are generated, and retrieval recommendation is conducted on the text information through a retrieval algorithm; according to the scheme, the security of the document information in the transmission and storage process is effectively guaranteed, and data leakage and tampering are prevented; the synchronization of a cloud local document directory and a file and a file version control technology are realized, so that a user can conveniently manage and update the document; the user intention can be accurately recognized, the accurate classification label is generated, the accurate retrieval recommendation service is provided, and the personalized requirement of the user is met.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of cloud document processing, and particularly relates to a knowledge intelligent classification label generation method and system for cloud documents. BACKGROUND

[0002] With the rapid development of information technology, a large amount of document materials are stored in the cloud. However, these documents lack effective classification and management, resulting in many difficulties for users in searching for required information. Traditional document classification methods often rely on manual annotation, which is not only inefficient but also prone to inaccurate classification. In addition, data security problems are increasingly prominent during document transmission and storage, and how to ensure the security of document information during transmission and storage has become a problem to be solved. Therefore, a method and system are needed that can automatically, efficiently and accurately classify and generate labels for cloud documents while ensuring data security. SUMMARY

[0003] The present application aims to solve the above problems, and designs a knowledge intelligent classification label generation method and system for cloud documents.

[0004] The first aspect of the present application provides a knowledge intelligent classification label generation method for cloud documents, which comprises the following steps:

[0005] The document information is transmitted by using a secure transmission encryption method, the cloud is transmitted to the local document, and a cloud local document database is established;

[0006] The cloud local document directory and files are synchronized by using a data synchronization and encoding method, and the cloud local document database is updated;

[0007] The text information is processed for information classification, classification labels are generated, and the text information is searched and recommended by using a search algorithm.

[0008] Further, in the knowledge intelligent classification label generation method for cloud documents, the cloud local document database is established by transmitting the document information by using a secure transmission encryption method, comprising:

[0009] The cloud file is obtained, and a secure file transmission protocol is used for transmission, which can be SFTP protocol and FTPS protocol;

[0010] A strong encryption algorithm is used to encrypt the transmitted data, so that only authorized users can decrypt and access;

[0011] End-to-end encryption is implemented throughout the data transmission process, and each link from the data source to the destination is encrypted to prevent data leakage by intermediate nodes;

[0012] A data verification mechanism is implemented during data transmission, and digital signature and hash verification techniques can be selected to ensure that data is not tampered with during transmission;

[0013] The cloud is transmitted to the local document, and a cloud local document database is established.

[0014] Further, in the above-mentioned knowledge intelligent classification tag generation method of cloud documents, the cloud local document directory and file are synchronized by using data synchronization and encoding methods, and the cloud local document database is updated, including:

[0015] A file sharing method based on the HTTP protocol is used to allow users to create, modify, and manage files on a Web server;

[0016] P2P technology is used to directly synchronize files between different platforms, reducing dependence on external servers and protecting data privacy;

[0017] The cloud local document directory and file are synchronized by using a file version comparison method, and the cloud local document database is updated.

[0018] Further, in the above-mentioned knowledge intelligent classification tag generation method of cloud documents, the text information is processed for information classification, and a classification tag is generated, and a retrieval algorithm is used to retrieve and recommend the text information, including:

[0019] User operation information is obtained, and context semantics and Query enhancement techniques are used to accurately identify user intent;

[0020] According to the knowledge of the power enterprise, a slicing strategy is adopted for information classification to generate classification tags;

[0021] The RAG vector retrieval and improved Rerank algorithm are combined to improve the accuracy of knowledge retrieval;

[0022] By recording the interactive history of user responses, combined with fine-tuning and Prompt engineering techniques, accurate content generation is achieved;

[0023] Relying on the recommendation algorithm, the user's preferences are captured to achieve personalized content generation.

[0024] Further, in the above-mentioned knowledge intelligent classification tag generation method of cloud documents, the RAG vector retrieval and improved Rerank algorithm are combined to improve the accuracy of knowledge retrieval, including:

[0025] An initial search is performed, user input query information is obtained, the query is encoded into a vector using the RAG vector retrieval technology, and the query vector is searched in the vector index to find the most similar candidate documents;

[0026] The candidate documents are input into the improved Rerank model together with the query information, and the model reorders the candidate documents according to the fused features and context information;

[0027] The ranked document list is returned to the user, and the user can select the required documents according to the ranking results.

[0028] Further, in the above-mentioned knowledge intelligent classification tag generation method of cloud documents, the user operation information is obtained, and the context semantics and Query enhancement technology are used to realize accurate identification of the user's intention, comprising:

[0029] The user operation information is obtained, and the natural language processing technology is used to extract the context semantic information of the text, identify the keywords with the highest weight in the power enterprise work scene, and provide a basis for intention classification;

[0030] The Query enhancement technology is combined to analyze and process the user's query request and extract the user's key intention;

[0031] The documents of the power enterprise are classified, and corresponding classification tags are generated, which can be classified according to equipment types, fault types and operation procedures.

[0032] The second aspect of the present application provides a knowledge intelligent classification tag generation system of cloud documents, characterized in that the knowledge intelligent classification tag generation system of cloud documents comprises a data transmission module, a data synchronization module and a classification retrieval module, wherein:

[0033] The data transmission module is used for transmitting document information in a secure transmission encryption mode, transmitting the cloud to the local document, and establishing a cloud local document database;

[0034] The data synchronization module is used for synchronizing the cloud local document directory and file by using a data synchronization and encoding mode, updating the cloud local document database;

[0035] The classification retrieval module is used for processing text information, classifying information, generating classification tags, and retrieving and recommending text information by using a retrieval algorithm.

[0036] Further, in the above-mentioned knowledge intelligent classification tag generation system of cloud documents, the data transmission module comprises an acquisition submodule, a file transmission submodule, a transmission encryption submodule, an end-to-end encryption submodule, a data verification submodule and a storage submodule, wherein:

[0037] The acquisition sub-module is used for acquiring a cloud file and transmitting the cloud file by using a secure file transmission protocol, which can be SFTP protocol and FTPS protocol;

[0038] The file transmission sub-module is used for encrypting the transmitted data by using a strong encryption algorithm, so that only authorized users can decrypt and access the data;

[0039] The transmission encryption sub-module is used for implementing end-to-end encryption in the whole process of data transmission, and each link from the data source to the destination is encrypted to prevent data leakage of intermediate nodes;

[0040] The end-to-end encryption sub-module is used for setting model parameters, specifically time step, LSTM unit number, activation function and optimizer, to obtain a load prediction model;

[0041] The data verification sub-module implements a data verification mechanism in the data transmission process, which can be selected to use digital signature and hash verification technology, to ensure that the data is not tampered with in the transmission process;

[0042] The storage sub-module is used for transmitting the cloud to a local document and establishing a cloud local document database.

[0043] The beneficial effects are that the document information is transmitted by using a secure transmission encryption method, the cloud is transmitted to a local document, and a cloud local document database is established; the cloud local document directory and files are synchronized by using a data synchronization and encoding method, and the cloud local document database is updated; the text information is processed for information classification, a classification label is generated, and a retrieval algorithm is used for retrieving and recommending the text information. The scheme effectively guarantees the security of the document information in the transmission and storage process by using a plurality of secure transmission encryption methods, prevents data leakage and tampering; the cloud local document directory and files are synchronized by using a data synchronization and encoding method, and a file version control technology is used, which facilitates the user to manage and update the document and improves the efficiency of the document management; the user's intention can be accurately identified, accurate classification labels are generated, and accurate retrieval recommendation services are provided to meet the individual needs of the user. BRIEF DESCRIPTION OF DRAWINGS

[0044] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments with reference made to the accompanying drawings. The drawings are for purposes of illustration only and are not considered as limiting the application.

[0045] Figure 1 An embodiment of a cloud document knowledge intelligent classification label generation method in the embodiments of the application is shown in the embodiment schematic diagram.

[0046] Figure 2 This is a first schematic diagram of a cloud-based intelligent knowledge classification tag generation method for documents according to an embodiment of the present invention;

[0047] Figure 3 This is a second schematic diagram of a cloud-based document knowledge intelligent classification tag generation method according to an embodiment of the present invention.

[0048] Figure 4 This is a third schematic diagram of a cloud-based document knowledge intelligent classification tag generation method in an embodiment of the present invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0050] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0051] The present invention will now be described in detail with reference to the accompanying drawings, such as... Figure 1 As shown, a method for generating knowledge intelligent classification tags for cloud documents includes the following steps:

[0052] like Figure 2 As shown, step 101 involves transmitting document information using a secure transmission encryption method, transferring the document from the cloud to the local document, and establishing a cloud-local document database.

[0053] Specifically, in this embodiment, cloud files are acquired using a secure file transfer protocol, such as SFTP or FTPS. Strong encryption algorithms are used to encrypt the transmitted data, ensuring that only authorized users can decrypt and access it. End-to-end encryption is implemented throughout the data transmission process, encrypting each step from the data source to the destination to prevent data leakage at intermediate nodes. A data verification mechanism is implemented during data transmission, such as digital signatures and hash verification techniques, to ensure that the data is not tampered with during transmission. Finally, cloud files are transferred to local documents, establishing a cloud-based local document database.

[0054] Mainstream protocols such as SFTP and AS2 ensure the confidentiality and integrity of files during transmission by using encryption technologies such as SSH and TLS. SFTP utilizes the SSH protocol for encrypted communication, supports multiple authentication methods, and is suitable for secure file sharing between enterprises and external partners. AS2 is commonly used for B2B messaging, providing not only encryption and signing functions but also the generation of message receipts to meet auditing requirements.

[0055] like Figure 3 As shown, step 102 involves using data synchronization and encoding methods to synchronize the cloud and local document directories and files, and updating the cloud and local document database.

[0056] Specifically, this embodiment adopts a file sharing method based on the HTTP protocol, allowing users to create, modify, and manage files on the web server; it utilizes P2P technology to directly synchronize files between different platforms, reducing dependence on external servers and protecting data privacy; and it employs a file version comparison method to synchronize the cloud-local document directory with the files, updating the cloud-local document database.

[0057] This solution enables the synchronization of documents and knowledge across different devices and operating systems. The knowledge management system regularly updates and iterates the knowledge base to ensure the timeliness and accuracy of the knowledge. By introducing artificial intelligence technology, the system can automatically identify outdated knowledge content and update it based on the latest data and information. Simultaneously, during the knowledge preservation process, the knowledge content undergoes rigorous verification and validation to ensure its reliability and effectiveness.

[0058] like Figure 4 As shown, step 103 involves processing the text information, classifying it, generating classification tags, and using a retrieval algorithm to retrieve and recommend text information.

[0059] Specifically, in this embodiment, user operation information is obtained, and contextual semantics and query enhancement technology are used to achieve accurate identification of user intent;

[0060] The system acquires user operation information, uses natural language processing technology to extract contextual semantic information from the text, identifies the keywords with the highest weight in the power company's work scenario, and provides a foundation for intent classification. Combined with query enhancement technology, it analyzes and processes user query requests to extract key user intents. The system then classifies the power company's documents and generates corresponding classification tags, which can be selected to be classified according to equipment type, fault type, or operating procedures.

[0061] Based on the knowledge of power companies, a slicing strategy is adopted to classify information and generate classification labels;

[0062] By combining RAG vector retrieval and the improved Rerank algorithm, the accuracy of knowledge retrieval can be improved.

[0063] The initial retrieval is performed to obtain the user's input query information. Using RAG vector retrieval technology, the query is encoded into a vector, and the candidate document most similar to the query vector is searched in the vector index.

[0064] The candidate documents are obtained and input together with the query information into the improved Rerank model. The model re-ranks the candidate documents based on the fused features and contextual information. The ranked list of documents is then returned to the user, who can select the document they need based on the ranking results.

[0065] By recording the interaction history of user responses and combining fine-tuning and Prompt engineering techniques, precise content generation can be achieved.

[0066] By leveraging recommendation algorithms, user preferences are captured to generate personalized content.

[0067] RAG (Retrievable Augmented Generation) vector retrieval is based on converting documents and queries into vectors in a vector space, and measuring the relevance of documents and queries by calculating the similarity between vectors.

[0068] RAG (Related Acyclic Graph) technology combines information retrieval and language generation models. It first retrieves relevant information from a knowledge base, then uses this information as context input to a language model to generate more accurate and relevant answers. Recommendation systems typically combine collaborative filtering and content-based recommendation to compensate for their respective shortcomings. Collaborative filtering makes recommendations by analyzing the similarity between users, combining user preferences and content features to achieve more accurate personalized recommendations. To improve the accuracy and robustness of recommendations, hybrid recommendation methods are widely used.

[0069] Deep learning-based Rerank models employ deep neural networks, such as BERT and Transformer, as Rerank models. They take the text content of queries and documents as input, learn the semantic relationships between them, and output a ranking score.

[0070] A cloud-based intelligent knowledge classification tag generation system for documents includes a data transmission module, a data synchronization module, and a classification retrieval module. The data transmission module transmits document information from the cloud to the local file using secure encryption, establishing a cloud-local document database. The data synchronization module synchronizes the cloud-local document directory and files using data synchronization and encoding methods, updating the cloud-local document database. The classification retrieval module processes text information, classifies it, generates classification tags, and uses retrieval algorithms to search and recommend text information.

[0071] In this embodiment, the data transmission module includes an acquisition submodule, a file transfer submodule, a transmission encryption submodule, an end-to-end encryption submodule, a data verification submodule, and a storage submodule. Specifically: the acquisition submodule acquires files from the cloud and transmits them using a secure file transfer protocol, such as SFTP or FTPS; the file transfer submodule encrypts the transmitted data using a strong encryption algorithm, ensuring that only authorized users can decrypt and access it; the transmission encryption submodule implements end-to-end encryption throughout the data transmission process, encrypting each step from the data source to the destination to prevent data leakage at intermediate nodes; the end-to-end encryption submodule sets model parameters, specifically the time step, the number of LSTM units, the activation function, and the optimizer, to obtain the load prediction model; the data verification submodule implements a data verification mechanism during data transmission, optionally using digital signatures and hash verification techniques to ensure that the data is not tampered with during transmission; and the storage submodule transmits cloud documents to local files, establishing a cloud-local document database.

[0072] By implementing the above solution, secure encryption is used for document information transmission, transferring documents from the cloud to the local machine to establish a cloud-local document database. Data synchronization and encoding methods are employed to synchronize the cloud-local document directory and files, updating the cloud-local document database. Text information is processed and categorized, generating category tags, and a retrieval algorithm is used for searching and recommending text information. This solution effectively ensures the security of document information during transmission and storage, preventing data leakage and tampering; it achieves synchronization of the cloud-local document directory and files, and incorporates file version control technology, facilitating document management and updates for users; it accurately identifies user intent, generates accurate category tags, and provides precise search and recommendation services to meet users' personalized needs.

[0073] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for generating knowledge intelligent classification tags for cloud-based documents, characterized in that, The method for generating knowledge intelligent classification tags for cloud-based documents includes the following steps: Document information is transmitted using secure encryption methods, transferring documents from the cloud to the local machine to establish a cloud-local document database; By employing data synchronization and encoding methods, the cloud-local document directory and files are synchronized, and the cloud-local document database is updated. The text information is processed and classified to generate category tags. A retrieval algorithm is then used to retrieve and recommend text information.

2. The method for generating knowledge intelligent classification tags for cloud-based documents as described in claim 1, characterized in that, The method of using secure transmission encryption to transmit document information, transferring documents from the cloud to the local machine, and establishing a cloud-local document database includes: Retrieve files from the cloud and transfer them using secure file transfer protocols, such as SFTP and FTPS. The transmitted data is encrypted using a strong encryption algorithm to ensure that only authorized users can decrypt and access it; End-to-end encryption is implemented throughout the entire data transmission process, with encryption performed at every stage from the data source to the destination to prevent data leakage at intermediate nodes. Implementing a data verification mechanism during data transmission can be achieved by using digital signatures and hash verification technologies to ensure that the data is not tampered with during transmission. Transfer documents from the cloud to the local machine to create a cloud-local document database.

3. The method for generating knowledge intelligent classification tags for cloud-based documents as described in claim 1, characterized in that, The method of synchronizing data and encoding to synchronize the cloud-local document directory and files, and updating the cloud-local document database includes: It uses an HTTP-based file-sharing method, allowing users to create, modify, and manage files on the web server; By using P2P technology to directly synchronize files between different platforms, dependence on external servers is reduced, and data privacy is protected. By using a file version comparison method, the cloud-local document directory and files are synchronized, and the cloud-local document database is updated.

4. The method for generating knowledge intelligent classification tags for cloud-based documents as described in claim 1, characterized in that, The process of processing text information, classifying it, generating classification tags, and using retrieval algorithms to search and recommend text information includes: By acquiring user operation information and utilizing contextual semantics and query enhancement techniques, accurate identification of user intent can be achieved. Based on the knowledge of power companies, a slicing strategy is adopted to classify information and generate classification labels; By combining RAG vector retrieval and the improved Rerank algorithm, the accuracy of knowledge retrieval can be improved. By recording the interaction history of user responses and combining fine-tuning and Prompt engineering techniques, precise content generation can be achieved. By leveraging recommendation algorithms, user preferences are captured to generate personalized content.

5. The method for generating knowledge intelligent classification tags for cloud-based documents as described in claim 4, characterized in that, The combination of RAG vector retrieval and the improved Rerank algorithm enhances the accuracy of knowledge retrieval, including: The initial retrieval is performed to obtain the user's input query information. Using RAG vector retrieval technology, the query is encoded into a vector, and the candidate document most similar to the query vector is searched in the vector index. The candidate documents are obtained and input together with the query information into the improved Rerank model. The model re-ranks the candidate documents based on the fused features and context information. The sorted list of documents is returned to the user, who can then select the document they need based on the sorting results.

6. The method for generating knowledge intelligent classification tags for cloud-based documents as described in claim 4, characterized in that, The process of obtaining user operation information and utilizing contextual semantics and query enhancement techniques to achieve accurate identification of user intent includes: By acquiring user operation information and using natural language processing technology to extract contextual semantic information from the text, we can identify the keywords with the highest weight in the work scenario of power companies and provide a basis for intent classification. By combining query enhancement technology, user query requests are analyzed and processed to extract the user's key intent; The document classification of power companies will generate corresponding classification tags. You can choose to classify them according to equipment type, fault type, or operating procedures.

7. A cloud-based intelligent knowledge classification tag generation system for documents, characterized in that, The cloud-based document knowledge intelligent classification tag generation system includes a data transmission module, a data synchronization module, and a classification retrieval module, wherein: The data transmission module is used to transmit document information using secure and encrypted transmission methods, transferring documents from the cloud to the local document database and establishing a cloud-local document database. The data synchronization module is used to synchronize the cloud-local document directory and files using data synchronization and encoding methods, and to update the cloud-local document database. The classification and retrieval module is used to process text information, classify it, generate classification tags, and use retrieval algorithms to search and recommend text information.

8. The cloud-based document knowledge intelligent classification tag generation system as described in claim 7, characterized in that, The data transmission module includes an acquisition submodule, a file transfer submodule, a transmission encryption submodule, an end-to-end encryption submodule, a data verification submodule, and a storage submodule, wherein: The Acquisition submodule is used to acquire files from the cloud and transmit them using a secure file transfer protocol, which can be either SFTP or FTPS. The file transfer submodule is used to encrypt the transmitted data using a strong encryption algorithm, ensuring that only authorized users can decrypt and access it; The transmission encryption submodule is used to implement end-to-end encryption throughout the entire data transmission process, encrypting each link from the data source to the destination to prevent data leakage at intermediate nodes; The end-to-end encryption submodule is used to set model parameters, specifically time step, number of LSTM units, activation function, and optimizer, to obtain the load prediction model. The data verification submodule implements a data verification mechanism during data transmission, and can be configured to use digital signature and hash verification technologies to ensure that the data has not been tampered with during transmission. The storage submodule is used to transfer documents from the cloud to the local machine and establish a cloud-local document database.

Images