A file management label intelligent indexing method and system, and electronic equipment

Abstract

The present invention provides an archive management tag intelligent indexing method and system, and an electronic device. The present invention trains an archive management tag intelligent indexing model based on target keywords input in sequence by a user, the number of occurrences in target archives found by the user ranked from large to small, words ranked in a preset top position, and archive management tag data during the archive management tag indexing process. After the model training is completed, the target keywords input in sequence by the user are input into the archive management tag intelligent indexing model, and the archive management tags output by the archive management tag intelligent indexing model are arranged at the front. The remaining archives are sorted from high to low according to the scores and output to the user. When the user attempts to extract or query specific archive data, the query results are optimized according to the user's historical query behavior or the correlation between data, and related and potentially more valuable information is quickly obtained, thereby greatly improving the efficiency and accuracy of the tag intelligent indexing.

Description

Technical Field

[0001] The present invention relates to the technical field of archive management scheme design, and in particular to an archive management tag intelligent indexing method and system, and electronic equipment. Background Art

[0002] With the rapid growth of digital information, archival data management has become an important challenge in the field of modern information technology. Traditional data storage and retrieval systems face many limitations when processing large-scale and diverse archival data.

[0003] Most existing systems lack effective tools for analyzing and mining the potential connections between user-entered keywords and archival data. They typically offer only simple keyword-based searches, failing to identify deeper connections and patterns between data. This limits users' ability to quickly discover target archives within the data. When users attempt to retrieve or query specific archival data, existing systems often fail to optimize query results based on historical user query behavior or inter-data correlations. This prevents users from quickly accessing relevant and potentially more valuable information.

[0004] Therefore, the existing technology needs to be further developed. Summary of the Invention

[0005] The purpose of the present invention is to overcome the above-mentioned technical deficiencies and provide a method and system for intelligent indexing of archive management tags, as well as an electronic device, to solve the problems existing in the prior art.

[0006] To achieve the above technical objectives, according to a first aspect of the present invention, the present invention provides a method for intelligently indexing archive management tags, comprising:

[0007] S100, receiving archival data from data sources of different types and formats, and constructing an archival database; in the current archival management tag indexing process, obtaining target keywords inputted by the user in sequence, assigning preset weights to each target keyword in sequence according to the input order of the target keywords and a preset weight list, calculating the number of occurrences of each target keyword for each archive in the archival database in sequence, scoring each archive in the archival database according to the number of occurrences of each target keyword and the preset weight of each target keyword, sorting each archive from high to low according to the score and outputting the scores to the user, performing human-computer interaction with the user to confirm whether the target archive is found, and if so, obtaining the archival management tag inputted by the user, retrieving the target archive according to the archival management tag inputted by the user, obtaining the vocabulary in the target archive that is ranked in descending order according to the number of occurrences and ranked in the top preset position, and recording them as a target vocabulary set, recording the target keywords, target vocabulary set, and archival management tag data inputted by the user in the current archival management tag indexing process as a label indexing process data set, collecting a preset number of label indexing process data sets, and dividing the collected label indexing process data sets into a training data set and a validation data set according to a first preset ratio;

[0008] S200: Using the collected training data set to train the archive management label intelligent indexing model, using the validation data set to optimize the archive management label intelligent indexing model, optimizing the model parameters, and completing the archive management label intelligent indexing model training;

[0009] S300. When the model training is completed, the target keywords entered by the user are obtained in sequence, and the target keywords entered by the user in sequence are input into the archive management label intelligent indexing model. The archive management labels output by the archive management label intelligent indexing model are arranged at the front, and the remaining archives are sorted from high to low according to the scores and output to the user.

[0010] Specifically, the human-computer interaction with the user to confirm whether the target file is found includes:

[0011] Output a prompt signal to the user to indicate whether the target file has been found, and obtain a feedback signal input by the user, wherein the feedback signal input by the user includes a feedback signal confirming that the target file has been found and a feedback signal confirming that the target file has not been found, and confirm whether the target file has been found based on the feedback signal input by the user.

[0012] Specifically, human-computer interaction with the user to confirm whether the target file is found also includes:

[0013] If the target file is not found, the target keywords inputted by the user are retrieved again and the process S100 is executed again.

[0014] Specifically, the method of using the collected training data set to train the archive management label intelligent indexing model, using the verification data set to optimize the archive management label intelligent indexing model, optimizing the model parameters, and completing the archive management label intelligent indexing model training includes:

[0015] The training data set is input into the preset network layer in batches for training. The preset network layer includes a Transformer network layer. The Transformer network layer is used to predict the next label indexing process data set based on the current label indexing process data set through forward propagation, thereby obtaining the predicted loss value; the loss value of the preset network layer is calculated and input into the optimizer for optimization to determine the direction in which the parameter gradient of the archive management label intelligent indexing model decreases fastest; the archive management label intelligent indexing model performs back propagation based on the loss value and the parameter gradient of the model to optimize the parameters of the archive management label intelligent indexing model.

[0016] Specifically, the method of using the collected training data set to train the archive management label intelligent indexing model, using the verification data set to optimize the archive management label intelligent indexing model, optimizing the model parameters, and completing the archive management label intelligent indexing model training includes:

[0017] After each training, the verification data set is input into the preset network layer of the previous training in batches for model parameter verification, and cyclic training is performed. The total number of training rounds is set to the first preset round; the loss value of the preset network layer is recorded, and it is determined whether the loss value meets the first preset condition. Based on the judgment result, it is determined whether to end the training and output the model parameters.

[0018] Specifically, determining whether the loss value satisfies the first preset condition, and determining whether to terminate the training and output the model parameters based on the determination result, includes:

[0019] If the loss value meets the first preset condition, the training is terminated and the parameters of the current archive management label intelligent indexing model are output; if the loss value does not meet the first preset condition, the training continues.

[0020] Specifically, the first preset condition includes:

[0021] After the second preset round of training, the loss values ​​obtained in the next round of training are greater than or equal to the loss values ​​that occurred during the second preset round of training.

[0022] Specifically, the scoring of each file in the file database according to the number of occurrences of each target keyword and the preset weight of each target keyword includes:

[0023] Select a file and select a target keyword from the file. Calculate the product of the preset weight of the target keyword and the number of times it appears in the file, and record it as the first score of the target keyword. Calculate the sum of the first scores of all target keywords in the file, and record it as the score of the file.

[0024] According to a second aspect of the present invention, there is provided a file management tag intelligent indexing system, comprising:

[0025] The acquisition module is used to receive archive data from data sources of different types and formats and build an archive database; it is used to obtain the target keywords entered by the user in sequence during the current archive management tag indexing process;

[0026] The control module is used to assign preset weights to each target keyword in accordance with the input order of the target keyword and the preset weight list, calculate the number of occurrences of each target keyword in each file in the file database in turn, score each file in the file database according to the number of occurrences of each target keyword and the preset weight of each target keyword, sort each file from high to low according to the score and output it to the user, conduct human-computer interaction with the user, confirm whether the target file is found, and if so, obtain the file management label input by the user, call the target file according to the file management label input by the user, obtain the words with the number of occurrences in the target file according to the ranking from large to small, and rank the words with the top preset ranking as the target word set, and in the process of indexing the file management label, the target keywords, target words input by the user in turn are sorted and output to the user, and the user is contacted with the target keyword in the file management label. The archive management label data is collected and recorded as a label indexing process data set, a preset number of label indexing process data sets are collected, and the collected label indexing process data sets are divided into a training data set and a verification data set according to a first preset ratio; the collected training data set is used to train the archive management label intelligent indexing model, and the verification data set is used to optimize the archive management label intelligent indexing model, optimize the model parameters, and complete the archive management label intelligent indexing model training; when the model training is completed, the target keywords entered by the user in sequence are obtained, the target keywords entered by the user in sequence are input into the archive management label intelligent indexing model, the archive management labels output by the archive management label intelligent indexing model are arranged in the front, and the remaining archives are sorted from high to low according to the scores and output to the user.

[0027] According to a third aspect of the present invention, there is provided an electronic device comprising: a memory; and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the above-mentioned intelligent indexing method for archive management tags is implemented.

[0028] Beneficial effects:

[0029] The present invention is based on the target keywords inputted in sequence by the user during the archive management tag indexing process, the number of occurrences in the target archives found by the user ranked from large to small, the words ranked in the top preset positions, and the archive management tag data of the target archives found by the user, to train an archive management tag intelligent indexing model. When the model training is completed, the target keywords inputted in sequence by the user are obtained, the target keywords inputted in sequence by the user are inputted into the archive management tag intelligent indexing model, the archive management tags outputted by the archive management tag intelligent indexing model are arranged at the front, and the remaining archives are sorted from high to low according to the scores and output to the user. When the user attempts to extract or query specific archive data, the query results are optimized according to the user's historical query behavior or the correlation between data, so as to quickly obtain related and potentially more valuable information, thereby greatly improving the efficiency, intelligence and accuracy of the archive management tag intelligent indexing. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of a method for intelligent indexing of archive management tags provided in a specific embodiment of the present invention;

[0031] Figure 2 This is a schematic diagram of the system composition of the file management tag intelligent indexing system provided in a specific embodiment of the present invention;

[0032] Figure 3 It is a structural diagram of the intelligent indexing model of archive management tags provided in a specific embodiment of the present invention. DETAILED DESCRIPTION

[0033] In order to enable those skilled in the art to better understand the technical solution of the present invention, the technical solution of the present invention is clearly and completely described below in conjunction with the drawings of the present invention. Based on the embodiments in this application, other similar embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of this application. In addition, the directional words mentioned in the following embodiments, such as "up", "down", "left", "right", etc., are only reference to the directions of the drawings. Therefore, the directional words used are used to illustrate rather than limit the invention.

[0034] The present invention will be further described below with reference to the accompanying drawings and preferred embodiments.

[0035] See also Figure 1 The present invention provides a method for intelligent indexing of archive management tags, comprising:

[0036] S100: Receive archival data from data sources of different types and formats and construct an archival database; in the current archival management tag indexing process, obtain target keywords inputted by the user in sequence, assign preset weights to each target keyword in sequence according to the input order of the target keywords and a preset weight list, calculate the number of occurrences of each target keyword for each archive in the archival database in sequence, score each archive in the archival database according to the number of occurrences of each target keyword and the preset weight of each target keyword, sort each archive from high to low according to the score and output it to the user, perform human-computer interaction with the user to confirm whether the target archive is found, and if so, obtain the archival management tag inputted by the user, retrieve the target archive according to the archival management tag inputted by the user, obtain the vocabulary in the target archive that is ranked in descending order by the number of occurrences and ranked in the top preset position, and record it as a target vocabulary set; record the target keywords, target vocabulary set, and archival management tag data inputted by the user in the current archival management tag indexing process as a label indexing process data set; collect a preset number of label indexing process data sets, and divide the collected label indexing process data sets into a training data set and a validation data set according to a first preset ratio.

[0037] Specifically, the human-computer interaction with the user to confirm whether the target file is found includes:

[0038] Output a prompt signal to the user to indicate whether the target file has been found, and obtain a feedback signal input by the user, wherein the feedback signal input by the user includes a feedback signal confirming that the target file has been found and a feedback signal confirming that the target file has not been found, and confirm whether the target file has been found based on the feedback signal input by the user.

[0039] Specifically, human-computer interaction with the user to confirm whether the target file is found also includes:

[0040] If the target file is not found, the target keywords input by the user are retrieved again and the process S100 is executed again, that is, the next file management tag index is performed.

[0041] It can be understood that the above process can realize the continuous updating of the data set of the archive management label intelligent indexing model of the present invention and ensure the reliability of the data set. After the training of the archive management label intelligent indexing model of the present invention is completed and put into use, the archive management label intelligent indexing model of the present invention can still be trained using the continuously updated data set to realize the continuous optimization of the archive management label intelligent indexing model of the present invention and continuously improve the efficiency and accuracy of the label intelligent index of the present invention.

[0042] Specifically, the scoring of each file in the file database according to the number of occurrences of each target keyword and the preset weight of each target keyword includes:

[0043] Select a file and select a target keyword from the file. Calculate the product of the preset weight of the target keyword and the number of times it appears in the file, and record it as the first score of the target keyword. Calculate the sum of the first scores of all target keywords in the file, and record it as the score of the file.

[0044] It should be noted that the steps before step S100 include:

[0045] A preset number, a first preset ratio, a first preset round, a second preset round, a preset weight list, and a first preset ranking are preset in the control module.

[0046] It can be understood that the preset number, first preset ratio, first preset round, second preset round, preset weight list, and previous preset ranking can be specifically set according to the actual needs of the user of the present invention. The present invention does not limit the specific values ​​of the above parameters, and there is no rule for setting the specific values ​​of the above parameters, as long as it is applicable to the intelligent indexing method of archive management tags proposed in the present invention.

[0047] Preferably, the present invention sets the first preset number to 1000, the present invention sets the first preset ratio to 8:2, the present invention sets the first preset round to 20,000 times, the present invention sets the second preset round to 300, the present invention sets the preset weight list to {100, 90, 80, 70, 60, 50, 40, 30, 20, 10, 0}, and the present invention sets the first preset ranking to the top 20. The above settings can further improve the performance of the model and the efficiency of model training, and greatly improve the accuracy and reliability of the prediction of the archive management label intelligent indexing model of the present invention.

[0048] It can be understood that when a user performs intelligent indexing of archive management tags, if the user finds the target archive by sequentially inputting target keywords, there is a certain coupling relationship between the target keywords sequentially input by the user and the words in the target archive whose occurrences are ranked from large to small and ranked in the top preset positions, that is, there is a certain coupling relationship between the target keywords sequentially input by the user and the content of the target archive. Based on this, the present invention trains the intelligent indexing model of archive management tags.

[0049] Specifically, the receiving of archival data from data sources of different types and formats and the building of an archival database include:

[0050] Receive archival data from data sources of different types and formats, support multiple data access protocols, and can handle large numbers of concurrent data streams.

[0051] Specifically, the system supports multiple data access protocols, including HTTP / HTTPS, FTP, SFTP, WebDAV, and various database connection protocols, such as JDBC and ODBC.

[0052] Specifically, a concurrent processing mechanism is used to handle a large number of concurrent data streams. The engine can optimize data transmission, reduce latency, and ensure the stability and efficiency of data access even under high load conditions.

[0053] Specifically, the receiving of archival data from data sources of different types and formats and the building of an archival database include:

[0054] Use OCR optical character recognition technology to scan paper files and obtain electronic file data.

[0055] Specifically, the data sources of different types and formats include:

[0056] File data source: including Excel data source and CSV data source, etc.

[0057] Database data sources: such as MySQL, MariaDB, Oracle, SQL Server, PostgreSQL, Greenplum, etc.

[0058] Big data engine data sources: such as Kylin, Cloudera Impala, Hive, MemSQL, Presto, MongoDB, etc.

[0059] Third-party interface data sources: such as OData, REST, and DataInsight.

[0060] S200. Use the collected training data set to train the archive management label intelligent indexing model, use the verification data set to optimize the archive management label intelligent indexing model, optimize the model parameters, and complete the archive management label intelligent indexing model training.

[0061] Specifically, the method of using the collected training data set to train the archive management label intelligent indexing model, using the verification data set to optimize the archive management label intelligent indexing model, optimizing the model parameters, and completing the archive management label intelligent indexing model training includes:

[0062] The training data set is input into the preset network layer in batches for training. The preset network layer includes a Transformer network layer. The Transformer network layer is used to predict the next label indexing process data set based on the current label indexing process data set through forward propagation, thereby obtaining the predicted loss value; the loss value of the preset network layer is calculated and input into the optimizer for optimization to determine the direction in which the parameter gradient of the archive management label intelligent indexing model decreases fastest; the archive management label intelligent indexing model performs back propagation based on the loss value and the parameter gradient of the model to optimize the parameters of the archive management label intelligent indexing model.

[0063] It can be understood that Transformer is a deep learning model based on the self-attention mechanism, which consists of a multi-head attention mechanism and a feedforward neural network. It can process sequence data and capture long-distance dependencies. The Transformer model does not contain traditional convolutional layers, but processes each element in the input sequence through an attention layer. The typical structure of Transformer includes an encoder and a decoder, where the encoder is composed of multiple identical layers stacked together, and the decoder adds a self-attention layer on the basis of the encoder to process sequence generation tasks. The present invention uses Transformer to supplement global context information, enhances the model's capabilities in feature extraction and global understanding, greatly improves the performance of the model, and greatly reduces the computational cost.

[0064] Specifically, the preset network layer also includes a ResNet network layer, which is used to classify and predict the output results based on the collected label index process data set, and the classification results are archive management labels.

[0065] It should be noted here that ResNet is a deep convolutional neural network that solves the gradient disappearance problem in deep network training by introducing residual connections. The core component of ResNet is the residual block, which usually contains two or three convolutional layers and a skip connection. The latter allows the gradient to bypass these convolutional layers directly. The network structure of ResNet can be very deep. These networks extract the features of the target keywords, target vocabulary sets, and archive management label data entered by the user in the archive management label indexing process by stacking multiple residual blocks, and classify and predict the output results through global average pooling and fully connected layers. The classification result is the archive management label.

[0066] See also Figure 3The present invention combines the ResNet network layer and the Transformer network layer. ResNet is good at extracting the local features of the target keywords, target vocabulary sets, and archive management label data entered by the user in sequence during the archive management label indexing process, while Transformer can supplement the global context information. This combination can enhance the model's capabilities in feature extraction and global understanding, greatly improve the model's performance, and greatly reduce the computational cost.

[0067] It can be understood that the ResNet network layer is used to classify and predict the output results through the label index process data set collected in the training data set and the validation data set. The specific steps are:

[0068] (1) First, the training dataset is placed in the Backbones network layer of the model. Backbone refers to a series of convolutional layers that constitute the skeleton of the neural network. The main function of these layers is to extract the features of the input data. The Backbone network is usually composed of multiple convolutional layers, pooling layers, and activation functions. It can extract meaningful feature representations from the original data. Backbones is composed of multiple Backbones, and the core network layer in Backbone is the Resnet network layer. After the dataset is input into the model, the label index process data set will be input into Backbones for training;

[0069] (2) When constructing Backbone, the Resnet network layer and archive storage location data were added. After the training data set was input, Resnet began to train layer by layer. The training extracted features from the label indexing process data set to assist Transformer in predicting archive management labels. The extracted features were used to obtain the location of the target archive with the help of the archive storage location, making it easier for archive managers to locate paper archives.

[0070] Understandably, the archive management tags include:

[0071] File name: including document title, document number, issuing unit, etc.;

[0072] File box numbering: Assign a unique number to each file box for easy search and management.

[0073] Specifically, the method of using the collected training data set to train the archive management label intelligent indexing model, using the verification data set to optimize the archive management label intelligent indexing model, optimizing the model parameters, and completing the archive management label intelligent indexing model training includes:

[0074] After each training, the verification data set is input into the preset network layer of the previous training in batches for model parameter verification, and cyclic training is performed, with the total number of training rounds set to the first preset round; the loss value of the preset network layer is recorded, and it is determined whether the loss value meets the first preset condition, and based on the judgment result, it is determined whether to end the training and output the model parameters.

[0075] Specifically, after each training, the verification data set is input into the preset network layer of the previous training in batches to verify the model parameters, and the cyclic training is performed, including the following methods:

[0076] When the number of training times is insufficient and a model needs to be optimized, the only option is to start training from the beginning. Therefore, the present invention develops a method for continuing the following training based on the previously trained model parameter file.

[0077] (1) Provide custom parameters, which can be set according to user needs to continue training based on any training parameters;

[0078] (2) Load the parameters of the custom model file;

[0079] (3) Put the validation data set into the custom parameter network layer for validation, obtain the loss value and record the model parameters of the model corresponding to the minimum loss value during the training process;

[0080] (4) Based on this set of model parameters, forward propagation training and back propagation are performed, and the optimizer optimizes the parameters to obtain a new set of data;

[0081] (5) Put the verification data into the new data network layer for inference to obtain the loss value;

[0082] (6) Perform cyclic training to obtain the best model and complete model optimization.

[0083] It is understood that the training method used in the present invention is batch training, which means that each time the model parameters are updated, only a portion of the samples in the validation dataset is used, called a batch. The advantages of batch training are that it can reduce memory consumption, speed up training, increase randomness, and facilitate model generalization.

[0084] It should be noted here that, in the present invention, batch_size=2, where batch_size is the batch size. For example, when batch_size=2, it means that two label index process data sets are selected from the verification data set each time and put into the model for verification.

[0085] Specifically, determining whether the loss value satisfies the first preset condition, and determining whether to terminate the training and output the model parameters based on the determination result, includes:

[0086] If the loss value meets the first preset condition, the training is terminated and the parameters of the current archive management label intelligent indexing model are output; if the loss value does not meet the first preset condition, the training is continued;

[0087] The first preset condition is that after the second preset round of training, the loss value obtained in the next round of training is greater than or equal to the loss value that occurred during the second preset round of training.

[0088] It can be understood that the present invention preferably sets the maximum round of cyclic training to 20,000 times, which can effectively ensure the accuracy of the model prediction of the present invention and record the loss value of the model training. The present invention sets the second preset round to 300, that is, when the loss value of the training no longer decreases within 300 rounds of training, the round with the smallest loss value in the 300 rounds of training is saved as the parameter of the best verification round, and the best model file is generated. The above setting makes the training rounds no longer rely on manual judgment, but uses the deep learning model for automatic training, which improves the accuracy of the model prediction, and automatically ends the training when the accuracy meets the requirements, which greatly saves training time and can effectively prevent overfitting, thereby greatly improving the intelligence level of the present invention and the efficiency of model training.

[0089] Furthermore, the specific process of training the intelligent indexing model for archive management tags of the present invention is as follows:

[0090] (1) Obtain the latent vector mean mu and logarithmic variance logvar in the forward propagation process of the archive management label intelligent indexing model;

[0091] (2) Calculate a reasonable loss value representing the training effect:

[0092] Calculate the divergence: k1 = -0.5*(1+logvar-mu^2-(e^logvar));

[0093] Calculate the loss between the label index process data set and the predicted value in the validation data set:

[0094] loss = L1 + k1 * k1_weight;

[0095] Among them, loss is the loss value, k1 is the divergence, k1_weight is the proportion of k1, and its value is 10. L1 is the absolute difference between the predicted value and the true value. The calculation formula of L1 is as follows:

[0096]

[0097] Among them, A is the L1 loss in pytorch, which calculates the average or sum of the absolute differences between two tensors; the shape of A is N*C*H*W, which indicates that it is a four-dimensional tensor, N represents the number of samples (Batch Size), indicating that there are N data points, C represents the number of channels (Channels), H and W represent height and width respectively, which are usually the two-dimensional spatial dimensions of the input features; P represents the padding value, which is a Boolean type, and the shape of P is c*h*w*1. The dimensions of P are the same as the dimensions of A in terms of channel, height, and width, but there is no number of samples N, which means that P is a mask or padding information for a feature map of each sample. The Boolean value of P may be used to mark whether the elements in A are filled (True) or valid (False).

[0098] (3) The loss value when using the validation dataset to validate the current model network;

[0099] (4) If the validation loss in this round is less than that in the previous round, the best model parameters are replaced with the training model of this round;

[0100] (5) Put the rounds and loss values ​​into the dictionary one by one;

[0101] (6) Compare the loss value of the current round with the loss value of the previous 300 rounds (parameters are provided and can be set according to actual needs, the default is 300), that is, traverse the loss values ​​from epoch-300 to epoch-1 (epoch is the training round, and comparison is performed when the epoch is greater than 300). If the loss values ​​of the current round of epoch are all greater than the loss values ​​from (epoch-200) to (epoch-1), then stop training.

[0102] (7) Save the best model parameters in the cyclic validation as a ckpt model file.

[0103] Specifically, determining whether the loss value satisfies the first preset condition, and determining whether to terminate the training and output the model parameters based on the determination result, includes:

[0104] If the loss value meets the first preset condition, the training is terminated and the parameters of the current archive management label intelligent indexing model are output; if the loss value does not meet the first preset condition, the training continues.

[0105] Specifically, the first preset condition includes:

[0106] After the second preset round of training, the loss values ​​obtained in the next round of training are greater than or equal to the loss values ​​that occurred during the second preset round of training.

[0107] S300. When the model training is completed, the target keywords entered by the user are obtained in sequence, and the target keywords entered by the user in sequence are input into the archive management label intelligent indexing model. The archive management labels output by the archive management label intelligent indexing model are arranged at the front, and the remaining archives are sorted from high to low according to the scores and output to the user.

[0108] It can be understood that the present invention is based on the target keywords input by the user in sequence during the archive management tag indexing process, the number of occurrences in the target archives found by the user ranked from large to small, the words ranked in the top preset positions, and the archive management tag data of the target archives found by the user, to train the archive management tag intelligent indexing model. When the model training is completed, the target keywords input by the user in sequence are obtained, and the target keywords input by the user in sequence are input into the archive management tag intelligent indexing model. The archive management tags output by the archive management tag intelligent indexing model are arranged at the front, and the remaining archives are sorted from high to low according to the scores and output to the user. When the user tries to extract or query specific archive data, the query results are optimized according to the user's historical query behavior or the correlation between data, so as to quickly obtain related and potentially more valuable information, thereby greatly improving the efficiency, intelligence and accuracy of the archive management tag intelligent indexing.

[0109] See also Figure 2 The present invention provides another embodiment, which provides an intelligent indexing system for archive management tags. The intelligent indexing system for archive management tags includes:

[0110] The acquisition module 100 is used to receive archive data from data sources of different types and formats and build an archive database; it is used to obtain target keywords entered by the user in sequence during the current archive management tag indexing process;

[0111] The control module 200 is used to assign preset weights to each target keyword in accordance with the input order of the target keyword and the preset weight list, calculate the number of occurrences of each target keyword in each file in the file database in turn, score each file in the file database according to the number of occurrences of each target keyword and the preset weight of each target keyword, sort each file from high to low according to the score and output it to the user, perform human-computer interaction with the user, confirm whether the target file is found, and if so, obtain the file management tag input by the user, call the target file according to the file management tag input by the user, obtain the words with the number of occurrences in the target file ranked from large to small, and record them as the target word set, and in the process of indexing the file management tag, the target keywords input by the user, the target words ... The vocabulary set and archive management label data are recorded as the label indexing process data set, a preset number of label indexing process data sets are collected, and the collected label indexing process data sets are divided into a training data set and a verification data set according to a first preset ratio; the system is used to train the archive management label intelligent indexing model using the collected training data set, and optimize the archive management label intelligent indexing model using the verification data set, optimize the model parameters, and complete the archive management label intelligent indexing model training; the system is used to obtain the target keywords input by the user in sequence after the model training is completed, input the target keywords input by the user in sequence into the archive management label intelligent indexing model, arrange the archive management labels output by the archive management label intelligent indexing model at the front, and sort the remaining archives from high to low according to the scores and output them to the user.

[0112] It should be noted here that the present invention is based on the target keywords input by the user in sequence during the archive management tag indexing process, the number of occurrences in the target archives found by the user ranked from large to small, the words ranked in the top preset positions, and the archive management tag data of the target archives found by the user, to train the archive management tag intelligent indexing model. When the model training is completed, the target keywords input by the user in sequence are obtained, and the target keywords input by the user in sequence are input into the archive management tag intelligent indexing model. The archive management tags output by the archive management tag intelligent indexing model are arranged at the front, and the remaining archives are sorted from high to low according to the scores and output to the user. When the user tries to extract or query specific archive data, the query results are optimized according to the user's historical query behavior or the correlation between data, so as to quickly obtain related and potentially more valuable information, thereby greatly improving the efficiency, intelligence and accuracy of the archive management tag intelligent indexing.

[0113] In a preferred embodiment, the present application further provides an electronic device, comprising:

[0114] A memory; and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the method for intelligent indexing of archive management tags is implemented. The computer device can be broadly defined as a server, a terminal, or any other electronic device with necessary computing and / or processing capabilities. In one embodiment, the computer device may include a processor, a memory, a network interface, a communication interface, etc. connected via a system bus. The processor of the computer device can be used to provide necessary computing, processing and / or control capabilities. The memory of the computer device may include a non-volatile storage medium and an internal memory. An operating system, a computer program, etc. may be stored in or on the non-volatile storage medium. The internal memory can provide an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface and the communication interface of the computer device can be used to connect and communicate with external devices via a network. When the computer program is executed by the processor, the steps of the method of the present invention are performed.

[0115] The present invention can be implemented as a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, causes the steps of the method of an embodiment of the present invention to be performed. In one embodiment, the computer program is distributed on a plurality of computer devices or processors coupled to a network so that the computer program is stored, accessed, and executed in a distributed manner by one or more computer devices or processors. A single method step / operation, or two or more method steps / operations, can be performed by a single computer device or processor or by two or more computer devices or processors. One or more method steps / operations can be performed by one or more computer devices or processors, and one or more other method steps / operations can be performed by one or more other computer devices or processors. One or more computer devices or processors can perform a single method step / operation, or perform two or more method steps / operations.

[0116] It will be understood by those skilled in the art that the method steps of the present invention can be performed by instructing relevant hardware such as a computer device or a processor through a computer program, and the computer program can be stored in a non-transitory computer-readable storage medium, which causes the steps of the present invention to be performed when the computer program is executed. Depending on the circumstances, any reference to memory, storage, database or other media herein may include non-volatile and / or volatile memory. Examples of non-volatile memory include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, magnetic tape, floppy disk, magneto-optical data storage device, optical data storage device, hard disk, solid-state disk, etc. Examples of volatile memory include random access memory (RAM), external cache memory, etc.

[0117] It can be understood that the present invention is based on the target keywords input by the user in sequence during the archive management tag indexing process, the number of occurrences in the target archives found by the user ranked from large to small, the words ranked in the top preset positions, and the archive management tag data of the target archives found by the user, to train the archive management tag intelligent indexing model. When the model training is completed, the target keywords input by the user in sequence are obtained, and the target keywords input by the user in sequence are input into the archive management tag intelligent indexing model. The archive management tags output by the archive management tag intelligent indexing model are arranged at the front, and the remaining archives are sorted from high to low according to the scores and output to the user. When the user tries to extract or query specific archive data, the query results are optimized according to the user's historical query behavior or the correlation between data, so as to quickly obtain related and potentially more valuable information, thereby greatly improving the efficiency, intelligence and accuracy of the archive management tag intelligent indexing.

[0118] The various technical features described above can be combined arbitrarily. Although not all possible combinations of these technical features are described, any combination of these technical features should be considered to be covered by this specification as long as such combination does not conflict.

[0119] The specific embodiments of the present invention described above do not limit the scope of protection of the present invention. Any other corresponding changes and modifications made based on the technical concept of the present invention should be included in the scope of protection of the claims of the present invention.

Claims

1. A method for intelligent indexing of archive management tags, characterized in that: The method comprises: S100, receiving archival data from data sources of different types and formats, and constructing an archival database; in the current archival management tag indexing process, obtaining target keywords inputted by the user in sequence, assigning preset weights to each target keyword in sequence according to the input order of the target keywords and a preset weight list, calculating the number of occurrences of each target keyword for each archive in the archival database in sequence, scoring each archive in the archival database according to the number of occurrences of each target keyword and the preset weight of each target keyword, sorting each archive from high to low according to the score and outputting the scores to the user, performing human-computer interaction with the user to confirm whether the target archive is found, and if so, obtaining the archival management tag inputted by the user, retrieving the target archive according to the archival management tag inputted by the user, obtaining the vocabulary in the target archive that is ranked in descending order according to the number of occurrences and ranked in the top preset position, and recording them as a target vocabulary set, recording the target keywords, target vocabulary set, and archival management tag data inputted by the user in the current archival management tag indexing process as a label indexing process data set, collecting a preset number of label indexing process data sets, and dividing the collected label indexing process data sets into a training data set and a validation data set according to a first preset ratio; S200: Using the collected training data set to train the archive management label intelligent indexing model, using the validation data set to optimize the archive management label intelligent indexing model, optimizing the model parameters, and completing the archive management label intelligent indexing model training; S300: After model training is completed, the target keywords entered by the user are obtained and input into the file management tag intelligent indexing model. The file management tags output by the file management tag intelligent indexing model are arranged at the front, and the remaining files are sorted from high to low according to the scores and output to the user; Specifically, the file management tags include: File name: including document title, document number, and issuing unit; ‌File box number‌: Assign a unique number to each file box for easy search and management‌.

2. The intelligent indexing method for archive management tags according to claim 1, characterized in that: The human-computer interaction with the user to confirm whether the target file is found includes: Output a prompt signal to the user to indicate whether the target file has been found, and obtain a feedback signal input by the user, wherein the feedback signal input by the user includes a feedback signal confirming that the target file has been found and a feedback signal confirming that the target file has not been found, and confirm whether the target file has been found based on the feedback signal input by the user.

3. The intelligent indexing method for file management tags according to claim 1, characterized in that: Interact with the user to confirm whether the target file is found, including: If the target file is not found, the target keywords inputted by the user are retrieved again and the process S100 is executed again.

4. The intelligent indexing method for archive management tags according to claim 1, characterized in that: The method of using the collected training data set to train the archive management label intelligent indexing model, using the verification data set to optimize the archive management label intelligent indexing model, optimizing the model parameters, and completing the archive management label intelligent indexing model training includes: The training data set is input into the preset network layer in batches for training. The preset network layer includes a Transformer network layer. The Transformer network layer is used to predict the next label indexing process data set based on the current label indexing process data set through forward propagation, thereby obtaining the predicted loss value; the loss value of the preset network layer is calculated and input into the optimizer for optimization to determine the direction in which the parameter gradient of the archive management label intelligent indexing model decreases fastest; the archive management label intelligent indexing model performs back propagation based on the loss value and the parameter gradient of the model to optimize the parameters of the archive management label intelligent indexing model.

5. The intelligent indexing method for archive management tags according to claim 1, characterized in that: The method of using the collected training data set to train the archive management label intelligent indexing model, using the verification data set to optimize the archive management label intelligent indexing model, optimizing the model parameters, and completing the archive management label intelligent indexing model training includes: After each training, the verification data set is input into the preset network layer of the previous training in batches for model parameter verification, and cyclic training is performed, with the total number of training rounds set to the first preset round; the loss value of the preset network layer is recorded, and it is determined whether the loss value meets the first preset condition, and based on the judgment result, it is determined whether to end the training and output the model parameters.

6. The intelligent indexing method for file management tags according to claim 5, characterized in that: The determining whether the loss value satisfies the first preset condition, and determining whether to terminate the training and output the model parameters according to the determination result, includes: If the loss value meets the first preset condition, the training is terminated and the parameters of the current archive management label intelligent indexing model are output; if the loss value does not meet the first preset condition, the training continues.

7. The intelligent indexing method for archive management tags according to claim 6, characterized in that: The first preset condition includes: After the second preset round of training, the loss values ​​obtained in the next round of training are greater than or equal to the loss values ​​that occurred during the second preset round of training.

8. The intelligent indexing method for archive management tags according to claim 1, characterized in that: Scoring each file in the file database according to the number of occurrences of each target keyword and the preset weight of each target keyword includes: Select a file and select a target keyword from the file. Calculate the product of the preset weight of the target keyword and the number of times it appears in the file, and record it as the first score of the target keyword. Calculate the sum of the first scores of all target keywords in the file, and record it as the score of the file.

9. An intelligent indexing system for archive management tags, characterized by: The intelligent indexing method for archive management tags according to any one of claims 1 to 8 comprises: The acquisition module is used to receive archive data from data sources of different types and formats and build an archive database; it is used to obtain the target keywords entered by the user in sequence during the current archive management tag indexing process; The control module is used to assign preset weights to each target keyword in accordance with the input order of the target keyword and the preset weight list, calculate the number of occurrences of each target keyword in each file in the file database in turn, score each file in the file database according to the number of occurrences of each target keyword and the preset weight of each target keyword, sort each file from high to low according to the score and output it to the user, conduct human-computer interaction with the user, confirm whether the target file is found, and if so, obtain the file management label input by the user, call the target file according to the file management label input by the user, obtain the words with the number of occurrences in the target file according to the ranking from large to small, and rank the words with the top preset ranking as the target word set, and in the process of indexing the file management label, the target keywords, target words input by the user in turn are sorted and output to the user, and the user is contacted with the target keyword in the file management label. The archive management label data is collected and recorded as a label indexing process data set, a preset number of label indexing process data sets are collected, and the collected label indexing process data sets are divided into a training data set and a verification data set according to a first preset ratio; the collected training data set is used to train the archive management label intelligent indexing model, and the verification data set is used to optimize the archive management label intelligent indexing model, optimize the model parameters, and complete the archive management label intelligent indexing model training; when the model training is completed, the target keywords entered by the user in sequence are obtained, the target keywords entered by the user in sequence are input into the archive management label intelligent indexing model, the archive management labels output by the archive management label intelligent indexing model are arranged in the front, and the remaining archives are sorted from high to low according to the scores and output to the user.

10. An electronic device, characterized in that: include: Memory; and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the intelligent indexing method for archive management tags according to any one of claims 1 to 8 is implemented.

Images