Intelligent electronic file classifying and filing method
Through the construction of intelligent classification models, combined with support vector machines and deep learning algorithms, the problems of low efficiency and low accuracy of electronic file classification and archiving are solved, automated classification and self-optimization are realized, adapting to the needs of different scenarios, and improving classification Accuracy and stability.
Patent Information
- Application Number
- CN202411180752.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-27
- Publication Date
- 2024-12-03
AI Technical Summary
The existing electronic file classification and archiving methods are inefficient and have low accuracy, ignore the correlation between multiple attributes of the file, lack the evaluation and optimization mechanism of classification results, and cannot adapt to the needs in different scenarios.
By building an intelligent classification model, combining support vector machine and deep learning algorithm, multiple algorithms are integrated to identify and classify files, and manual classification learning optimization mechanism is adopted to collect user feedback to continuously improve the accuracy and adaptability of automatic classification, and to integrate different business units Classification standards to form a unified classification system.
It improves the accuracy and work efficiency of electronic file classification, reduces manual intervention, and the system can learn and optimize itself, adapt to the rapidly growing number of files, and ensures the stability and reliability of classification results.
Smart Images

Figure CN119066505A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of information technology document management, and in particular to an electronic file intelligent classification and archiving method. Background Art
[0002] With the advent of the digital age, electronic files are increasingly used in various fields, and the speed and quantity of their generation are also growing exponentially. However, with the rapid increase in the number of electronic files, how to efficiently and accurately classify and archive these files has become an urgent problem to be solved. Traditional electronic file classification and archiving methods usually rely on manual operations, which is not only inefficient but also prone to errors. Since the attribute information of electronic files is complex and diverse, how to effectively extract this information and classify and archive it is also a major challenge.
[0003] Existing electronic document classification and archiving methods usually only consider a single attribute or feature of a document for classification, while ignoring the correlation between multiple attributes of a document. This may lead to inaccurate and one-sided classification results. At the same time, existing methods also lack an evaluation and optimization mechanism for classification results, and cannot be adjusted and improved according to the classification effect.
[0004] In summary, the existing electronic file classification and filing methods still have some problems, such as low efficiency, low accuracy, lack of flexibility, etc. Therefore, there is an urgent need for a method that can automatically and intelligently classify and file electronic files to improve work efficiency and accuracy and meet the classification and filing needs in different scenarios. Summary of the invention
[0005] The main purpose of the present invention is to provide an electronic file intelligent classification and archiving method, which can solve the problems raised in the above background technology.
[0006] To achieve the above-mentioned purpose, the electronic file intelligent classification and archiving method proposed in the present invention realizes the automatic classification of the electronic files by constructing an intelligent classification model and integrating the archive classification schemes, file material archiving scopes and archive retention period tables of each business unit.
[0007] Preferably, the intelligent classification model adopts a fusion of support vector machine and deep learning algorithm. By fusing multiple algorithms, the system can more accurately identify and classify files, reducing the error rate caused by manual classification.
[0008] Preferably, the intelligent classification model performs manual classification learning optimization on unclassified data and collects classification feedback from users to continuously improve the scope and accuracy of automatic classification.
[0009] Preferably, the manual classification learning mechanism includes the following steps: S1: The system identifies unclassified files and prompts the user to manually classify them; S2: After the user classifies these files, the system feeds this information back to the intelligent classification model; S3: re-learning and optimization.
[0010] Through the manual classification learning optimization mechanism, the model can gradually adapt to new file types and classification requirements, and the system can continuously retrain and optimize based on user feedback to improve the adaptability and classification effect of the model. This self-learning ability enables the system to continuously improve over time.
[0011] Preferably, the core of the method is to integrate the classification standards of different business units to ensure the uniformity and accuracy of the classification results.
[0012] Preferably, the method effectively integrates the classification schemes of each unit through a fusion algorithm of support vector machine and deep learning algorithm to form a comprehensive classification system. SVM is a classification method based on statistical learning theory, which can find an optimal decision boundary (hyperplane) in high-dimensional space to maximize the interval between different categories. Deep learning algorithms such as convolutional neural networks (CNN) and recurrent neural networks (RNN) can automatically extract hierarchical feature representations from raw data through multi-layer nonlinear transformations. These feature representations are very effective for complex data structures and pattern recognition tasks, but usually require a large amount of training data and computing resources.
[0013] Preferably, the electronic file intelligent classification and archiving method includes the following technical solution steps: S1: Demand analysis and data collection, Demand analysis: Communicate with each business unit to understand the specific requirements of their archive classification plan, document material archiving scope, and archive retention period schedule.
[0014] Data collection: Collect historical electronic file data from each unit, including file content, metadata (such as creation time, author, file type, etc.) and existing classification labels.
[0015] S2: Build intelligent classification model, Feature extraction: Perform feature extraction on the collected electronic files to extract key features (such as text content, keywords, file format, etc.), and store the extracted features as feature vectors for subsequent model training.
[0016] Support Vector Machine (SVM) model training: Use labeled historical data to train the SVM model and optimize model parameters to improve classification accuracy.
[0017] Deep learning model training: Choose a suitable deep learning architecture (such as convolutional neural network CNN or recurrent neural network RNN) and use the same feature vector for training.
[0018] Model Fusion: The outputs of the SVM model and the deep learning model are fused, and a comprehensive classification model is formed using methods such as weighted averaging or voting mechanisms.
[0019] S3: Automatic classification, Business data input: Input the electronic documents to be classified into the intelligent classification model.
[0020] Automatic classification execution: The model automatically classifies files based on the extracted features and the fused classification scheme, and generates classification results. The automatic classification mechanism can quickly process a large number of electronic files, reduce the need for manual intervention, and thus improve work efficiency. The system can complete the classification and archiving of files in a short period of time to adapt to the rapidly growing number of electronic files.
[0021] S4: Manual classification learning optimization, unclassified file identification: The system identifies unclassified or inaccurately classified files and prompts the user to perform manual classification.
[0022] User feedback collection: Users manually classify unclassified files, and the system records the user's classification results.
[0023] Model retraining: Use the classification results of user feedback as new training data to regularly retrain and optimize the intelligent classification model to improve classification accuracy and adaptability.
[0024] S5: Classification result output and optimization, Classification result evaluation: Regularly evaluate the classification results and use indicators such as accuracy and recall to measure the classification effect.
[0025] Optimization mechanism implementation: According to the evaluation results, adjust the model parameters or classification scheme to continuously optimize the classification effect. Regular classification result evaluation and optimization mechanism ensure that the classification effect of the system is always at a high level and that problems in classification can be discovered and corrected in a timely manner.
[0026] Preferably, the method significantly improves the accuracy of electronic file classification by combining a support vector machine (SVM) and a deep learning algorithm. The fusion model can better handle complex file features and reduce misclassification.
[0027] Preferably, the electronic file content adopts a feature extraction method to convert the electronic file content into a feature vector that can be used for model training. Different electronic files have different formats and attributes. Through precise feature extraction, these different types of files can be converted into a unified feature vector representation, so that the model can handle multiple types of files, enhancing the flexibility and adaptability of the system. The model can understand the file content more accurately, thereby making more accurate classification decisions.
[0028] Preferably, the method integrates the outputs of different models through a fusion algorithm (such as weighted average or voting mechanism) to ensure the consistency and accuracy of the classification results. By fusing the outputs of different models (such as support vector machine SVM and deep learning model), the advantages of each model can be fully utilized and complement each other, thereby improving the accuracy of classification. A single model may perform poorly on certain types of files, but a fusion model can reduce the possibility of misclassification by integrating the prediction results of different models and reduce the fluctuation of the overall classification performance caused by the instability or errors of a single model.
[0029] The present invention provides a method for intelligent classification and archiving of electronic files. It has the following beneficial effects: (1) The electronic file intelligent classification and archiving method extracts features from electronic files through the combination of vector machines and deep learning algorithms, extracts key features, and converts these features into feature vectors that can be used for model training. The extracted feature vectors are trained using support vector machines and deep learning algorithms respectively. The outputs of the SVM model and the deep learning model are fused, and then the weighted sum is calculated. By fusing the prediction results of different models, the misclassification that may be caused by a single model can be reduced, thereby improving the accuracy of classification, making the entire classification system more stable and reliable, and having better robustness to noise and abnormal data.
[0030] (2) This electronic file intelligent classification and archiving method integrates the archive classification schemes, file material archiving scopes, and archive retention schedules of various business units through the fusion of multi-source information to form a unified classification standard. The fusion algorithm can reduce the fluctuation of the overall classification performance caused by the instability or errors of a single model. Even if a model performs poorly in a specific situation, the correct predictions of other models can be adopted by the fusion algorithm, thereby ensuring the stability and reliability of the overall classification results.
[0031] (3) The electronic file intelligent classification and archiving method is learned through manual classification. Manual classification is a direct feedback to the automatic classification results. When the user finds that the automatic classification is inaccurate, he can correct it through manual classification. The corrected data will be used as a new training sample and re-entered into the model for training, allowing the system to continuously obtain new classification samples from users and conduct self-learning and optimization. This self-learning ability enables the system to continuously improve over time and always maintain a high classification performance. BRIEF DESCRIPTION OF THE DRAWINGS
[0032] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 the structures shown in these drawings without paying creative work.
[0033] Figure 1 It is a schematic diagram of the technical solution flow of the present invention; Figure 2 This is a partial process diagram of the technical solution of the present invention Figure 1 ; Figure 3 This is a partial process diagram of the technical solution of the present invention Figure 2 ; Figure 4 This is a partial process diagram of the technical solution of the present invention Figure 3 ; Figure 5 This is a partial process diagram of the technical solution of the present invention Figure 4 ; Figure 6 This is a partial process diagram of the technical solution of the present invention Figure 5 ; Figure 7 This is a schematic diagram of the manual classification process structure of the present invention.
[0034] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION
[0035] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0036] See also Figure 1-7The present invention proposes an intelligent classification and archiving method for electronic files. By constructing an intelligent classification model, the archive classification schemes and file information of each business unit are integrated to perform more accurate classification. The electronic file content is converted into a feature vector that can be used for model training. Through accurate feature extraction, these different types of files can be converted into a unified feature vector representation. The feature vector, as the input of the model, has a fixed dimension and format, which makes the model training process more standardized and controllable. At the same time, the classification results based on the feature vector are also easier to evaluate and optimize. The classification effect can be continuously improved by adjusting the model parameters or improving the feature extraction method, so that the model can handle multiple types of files, enhance the flexibility and adaptability of the system, and the model can understand the file content more accurately, so as to make more accurate classification decisions. By combining the algorithm model and the business "three-in-one" specification, the intelligent classification and archiving of electronic files are realized, which solves the problems of low efficiency and easy errors in traditional methods. At the same time, through feature extraction and model training, the classification is made more accurate and flexible.
[0037] In the present invention, when used, the technical solution is implemented as follows: S1: Demand analysis and data collection, Demand analysis: Communicate with each business unit to understand the specific requirements of their archive classification plan, document material archiving scope, and archive retention period schedule.
[0038] Data collection: Collect historical electronic file data from each unit, including file content, metadata (such as creation time, author, file type, etc.) and existing classification labels.
[0039] S2: Build intelligent classification model, Feature extraction: Perform feature extraction on the collected electronic files to extract key features (such as text content, keywords, file format, etc.), and store the extracted features as feature vectors for subsequent model training.
[0040] Support Vector Machine (SVM) model training: Use labeled historical data to train the SVM model and optimize model parameters to improve classification accuracy.
[0041] Deep learning model training: Choose a suitable deep learning architecture (such as convolutional neural network CNN or recurrent neural network RNN) and use the same feature vector for training.
[0042] Model Fusion: The outputs of the SVM model and the deep learning model are fused, and a comprehensive classification model is formed using methods such as weighted averaging or voting mechanisms.
[0043] S3: Automatic classification, Business data input: Input the electronic documents to be classified into the intelligent classification model.
[0044] Automatic classification execution: The model automatically classifies files based on the extracted features and fused classification schemes and generates classification results.
[0045] S4: Manual classification learning optimization, unclassified file identification: The system identifies unclassified or inaccurately classified files and prompts the user to perform manual classification.
[0046] User feedback collection: Users manually classify unclassified files, and the system records the user's classification results.
[0047] Model retraining: Use the classification results of user feedback as new training data to regularly retrain and optimize the intelligent classification model to improve classification accuracy and adaptability.
[0048] S5: Classification result output and optimization, Classification result evaluation: Regularly evaluate the classification results and use indicators such as accuracy and recall to measure the classification effect.
[0049] Implementation of optimization mechanism: According to the evaluation results, adjust the model parameters or classification scheme to continuously optimize the classification effect.
[0050] The system can be customized according to the needs of different business units, supports various types of file classification, and has good adaptability.
[0051] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. All equivalent structural changes made by using the contents of the present invention specification and drawings under the inventive concept of the present invention, or directly / indirectly applied in other related technical fields are included in the patent protection scope of the present invention.
Claims
1. A method for intelligent classification and archiving of electronic files, characterized by: By constructing an intelligent classification model, the archive classification schemes of each business unit, the filing scope of document materials and the archive retention period table are integrated to achieve automatic classification of the electronic files.
2. The method for intelligent classification and archiving of electronic files according to claim 1, characterized in that: The intelligent classification model adopts the integration of support vector machine and deep learning algorithm.
3. The method for intelligent classification and archiving of electronic files according to claim 2, characterized in that: The intelligent classification model performs manual classification learning optimization on the unclassified data and collects classification feedback from users.
4. The method for intelligent classification and archiving of electronic files according to claim 3, characterized in that: The mechanism of manual classification learning The following steps are involved: S1: The system identifies unclassified files and prompts the user to manually classify them; S2: After the user classifies these files, the system feeds this information back to the intelligent classification model; S3: Re-learning and optimization are performed.
5. The method for intelligent classification and archiving of electronic files according to claim 1, characterized in that: The core of the method is to integrate the classification standards of different business units to ensure the uniformity and accuracy of the classification results.
6. A method for intelligent classification and archiving of electronic files according to any one of claims 2 to 5, characterized in that: The method effectively integrates the classification schemes of various units through the fusion algorithm of support vector machine and deep learning algorithm to form a comprehensive classification system.
7. The method for intelligent classification and archiving of electronic files according to claim 1, characterized in that: The electronic file intelligent classification and archiving method includes the following technical solution steps: S1: demand analysis and data collection, demand analysis: communicate with each business unit to understand its archive classification plan, document material archiving scope and specific requirements of the archive retention period table. Data collection: collect historical electronic file data of each unit, including file content, metadata (such as creation time, author, file type, etc.) and existing classification labels. S2: build an intelligent classification model, feature extraction: extract features from the collected electronic files, extract key features (such as text content, keywords, file format, etc.), and store the extracted features as feature vectors for subsequent model training. Support vector machine (SVM) model training: use the annotated historical data to train the SVM model and optimize the model parameters to improve the classification accuracy. Deep learning model training: select a suitable deep learning architecture (such as convolutional neural network CNN or recurrent neural network RNN) and use the same feature vector for training. Model fusion: fuse the outputs of the SVM model and the deep learning model, and use weighted average or voting mechanism to form a comprehensive classification model. S3: Automatic classification, business data input: Input the electronic files to be classified into the intelligent classification model. Automatic classification execution: The model automatically classifies the files based on the extracted features and the fused classification scheme, and generates classification results. S4: Manual classification learning optimization, unclassified file identification: The system identifies unclassified or inaccurately classified files, and prompts the user to manually classify them. User feedback collection: The user manually classifies the unclassified files, and the system records the user's classification results. Model retraining: Use the classification results fed back by the user as new training data, and regularly retrain and optimize the intelligent classification model to improve classification accuracy and adaptability. S5: Classification result output and optimization, classification result evaluation: Regularly evaluate the classification results, and use indicators such as accuracy and recall rate to measure the classification effect. Optimization mechanism implementation: According to the evaluation results, adjust the model parameters or classification scheme to continuously optimize the classification effect.
8. The method for intelligent classification and archiving of electronic files according to claim 7, characterized in that: The method significantly improves the accuracy of electronic document classification by combining a support vector machine (SVM) and a deep learning algorithm.
9. The method for intelligent classification and archiving of electronic files according to claim 1, characterized in that: The electronic document content adopts a feature extraction method to convert the electronic document content into a feature vector that can be used for model training.
10. The method for intelligent classification and archiving of electronic files according to claim 2, characterized in that: The method combines the outputs of different models through a fusion algorithm (such as weighted average or voting mechanism) to ensure the consistency and accuracy of the classification results.
Citation Information
Patent Citations
A method and a system for automatic evaluation of digital files
CA2581466A1
Object-based OBIA-SVM-CNN remote sensing image classification method
CN109740631A
Image classification method based on deep learning and support vector machine fusion
CN114898151A
Electronic archive information intelligent management system
CN117556112A
Personnel archive digital automatic classification method and system based on vocabulary statistics
CN117851869A
Cited By
Electronic file intelligent sorting and filing method and system based on AI large model
CN119988614A
File management method and device, computer equipment, storage medium and program product
CN120763124A
