File migration method, client, and computer-readable storage medium

Abstract

Embodiments of the present application disclose a file migration method, a client and a computer readable storage medium, which are used for automatically migrating files. The method of the embodiments of the present application comprises the following steps: obtaining a current date and a client disk usage space corresponding to the current date; inputting the current date into a pre-trained client disk usage rate regression model and a cloud disk usage rate regression model respectively to obtain a client disk usage rate corresponding to the current date and a cloud disk usage rate corresponding to the current date; if the client disk usage space, the client disk usage rate corresponding to the current date and the cloud disk usage rate corresponding to the current date satisfy a preset condition, obtaining basic information of each file to be migrated in the client disk and basic operation behavior information of a user on the current date; determining a target migration file in the client disk according to the basic information of each file to be migrated and the basic operation behavior information of the user; and migrating the target migration file to a cloud disk.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the field of file migration, and more particularly to a file migration method, a client and a computer readable storage medium. BACKGROUND

[0002] With the use of the client by the user, the files in the client disk will be more and more, when the files stored in the client disk are more and more, the storage space occupied will be more and more, and the remaining available storage space will be less and less, so that the storage space of the client is not enough. In order to ensure that the client has enough remaining available storage space, the files that are not frequently used in the client disk need to be migrated to the cloud disk.

[0003] The existing file migration method is that the user selects the files to be migrated, and the client receives the migration instruction of the files to be migrated, and migrates the files to be migrated to the cloud disk. However, this method requires the user to actively select the files to be migrated, and if the user does not select the files to be migrated, the files in the client disk will be more and more, the storage space occupied will be more and more, and the remaining available storage space will be less and less, so that the storage space of the client is not enough. The user experience of using the client is poor. SUMMARY

[0004] Embodiments of the present application provide a file migration method, a client and a computer readable storage medium, which can automatically migrate files.

[0005] In a first aspect, the embodiments of the present application provide a file migration method, comprising:

[0006] obtaining a current date and a client disk usage space corresponding to the current date;

[0007] inputting the current date into a pre-trained client disk usage rate regression model to obtain a client disk usage rate corresponding to the current date;

[0008] inputting the current date into a pre-trained cloud disk usage rate regression model to obtain a cloud disk usage rate corresponding to the current date;

[0009] if the client disk usage space, the client disk usage rate corresponding to the current date and the cloud disk usage rate corresponding to the current date satisfy a preset condition, obtaining basic information of each file to be migrated in the client disk of the current date and basic operation behavior information of the user;

[0010] determining a target migration file in the client disk according to the basic information of each file to be migrated and the basic operation behavior information of the user;

[0011] migrate the target migration file to a cloud disk.

[0012] Optionally, before the inputting the current date into the pre-trained client disk usage rate regression model to obtain the client disk usage rate corresponding to the current date, the method further comprises:

[0013] obtaining date samples of a preset time period, the date samples being labeled with client disk usage rates corresponding to the date samples;

[0014] inputting the date samples into a client disk usage rate regression model to obtain predicted client disk usage rates output by the client disk usage rate regression model;

[0015] calculating, according to a client disk usage rate regression loss function, a client disk usage rate loss between the predicted client disk usage rates and the labeled client disk usage rates, and obtaining a trained client disk usage rate regression model when the client disk usage rate loss satisfies a first preset convergence condition.

[0016] Optionally, before the inputting the current date into the pre-trained cloud disk usage rate regression model to obtain the cloud disk usage rate corresponding to the current date, the method further comprises:

[0017] obtaining date samples of a preset time period, the date samples being labeled with cloud disk usage rates corresponding to the date samples;

[0018] inputting the date samples into a cloud disk usage rate regression model to obtain predicted cloud disk usage rates output by the cloud disk usage rate regression model;

[0019] calculating, according to a cloud disk usage rate regression loss function, a cloud disk usage rate loss between the predicted cloud disk usage rates and the labeled cloud disk usage rates, and obtaining a trained cloud disk usage rate regression model when the cloud disk usage rate loss satisfies a second preset convergence condition.

[0020] Optionally, the preset condition is that the client disk usage space is greater than or equal to a preset usage space threshold, the client disk usage rate is greater than or equal to a preset client disk usage rate threshold, and the cloud disk usage rate is less than or equal to a preset cloud disk usage rate threshold.

[0021] Optionally, before the obtaining the basic information of each to-be-migrated file in the client disk and the basic operation behavior information of the user on the current date, the method further comprises:

[0022] scanning the client disk to obtain information about whether each file in the client disk is to be migrated;

[0023] Determine the files to be migrated in the client disk according to the information of whether each file of the client disk is migrated.

[0024] Optionally, the method further comprises:

[0025] Sort the each file to be migrated in the order from the earliest to the latest according to the file creation time or the file modification time, and filter the first pre-set number of files to be migrated in the front of the order to obtain a first filtered set of files to be migrated;

[0026] Input the basic information of each file to be migrated in the first filtered set of files to be migrated and the basic operation behavior information of the user into a pre-trained logistic regression classification model to obtain the probability of not being used of each file to be migrated in the first filtered set of files to be migrated;

[0027] Sort each file to be migrated in the first filtered set of files to be migrated in the order from large to small according to the probability, and filter the second pre-set number of files to be migrated in the front of the order to obtain a second filtered set of files to be migrated;

[0028] Sort each file to be migrated in the second filtered set of files to be migrated in the order from large to small according to the file size, and filter the third pre-set number of files to be migrated to obtain the target file to be migrated.

[0029] Optionally, before the basic information of each file to be migrated in the first filtered set of files to be migrated and the basic operation behavior information of the user are input into the pre-trained logistic regression classification model, the method further comprises:

[0030] Obtain the basic information sample of each file to be migrated sample and the basic operation behavior information sample of the user of each date sample in a pre-set time period; the basic information sample of each file to be migrated includes file save path, file name, file size, file creation time or file modification time, and file opening times, etc., the basic operation behavior information of the user includes the application use list of the user and the application use duration of the user using each application, etc., the file opening times greater than 0 is a negative label, and equal to 0 is a positive label;

[0031] Calculate a probability of not being used of each file sample to be migrated according to a file opening number of each file sample to be migrated, and mark the probability in a basic information sample of each file sample to be migrated;

[0032] Input the basic information sample of each file sample to be migrated and the basic operation behavior information sample of the user into a logistic regression classification model, to obtain a predicted probability of not being used of each file sample to be migrated output by the logistic regression classification model;

[0033] Calculate a probability loss between the predicted probability and the marked probability according to a probability regression loss function, and when the probability loss meets a third preset convergence condition, obtain the trained logistic regression classification model.

[0034] In a second aspect, an embodiment of the present application provides a client, comprising:

[0035] An obtaining unit is configured to obtain a current date and a client disk usage space corresponding to the current date;

[0036] An input unit is configured to input the current date into a pre-trained client disk usage rate regression model to obtain a client disk usage rate corresponding to the current date;

[0037] The input unit is further configured to input the current date into a pre-trained cloud disk usage rate regression model to obtain a cloud disk usage rate corresponding to the current date;

[0038] A judgment unit is configured to, if the client disk usage space, the client disk usage rate corresponding to the current date, and the cloud disk usage rate corresponding to the current date meet a preset condition, obtain basic information of each file to be migrated in a client disk of the current date and basic operation behavior information of a user;

[0039] A determination unit is configured to determine a target migration file in the client disk according to the basic information of each file to be migrated and the basic operation behavior information of the user;

[0040] A migration unit is configured to migrate the target migration file to a cloud disk.

[0041] In a third aspect, an embodiment of the present application provides a client, comprising:

[0042] A central processing unit, a memory, an input and output interface, a wired or wireless network interface, and a power supply;

[0043] The memory is a transitory storage memory or a persistent storage memory;

[0044] The central processing unit is configured to communicate with the memory, and perform instruction operations in the memory to execute the foregoing file migration method.

[0045] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, which includes instructions, when the instructions are executed on a computer, causing the computer to perform the file migration method.

[0046] In a fifth aspect, an embodiment of the present application provides a computer program product including instructions, when the computer program product is executed on a computer, causing the computer to perform the file migration method.

[0047] From the above technical solutions, it can be seen that the embodiments of the present application have the following advantages: the client can be determined in real time whether to meet the preset conditions for file migration, and the target file to be migrated can be determined as long as the preset conditions are met, and the target file to be migrated is migrated to the cloud disk, the entire file migration process is automatically executed by the client, the user does not perceive, the file on the client disk can be guaranteed not to be too much, the client can be guaranteed to have enough remaining available storage space, the situation that the storage space of the client is not enough is reduced, and the user experience of using the client is improved. BRIEF DESCRIPTION OF DRAWINGS

[0048] Figure 1 A flowchart of a file migration method disclosed by an embodiment of the present application is shown in the figure;

[0049] Figure 2 A flowchart of a method for determining a target migration file disclosed by an embodiment of the present application is shown in the figure;

[0050] Figure 3 A structure diagram of a client disclosed by an embodiment of the present application is shown in the figure;

[0051] Figure 4 A structure diagram of another client disclosed by an embodiment of the present application is shown in the figure;

[0052] Figure 5 A structure diagram of another client disclosed by an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0053] The embodiments of the present application provide a file migration method, a client and a computer readable storage medium, which are used for automatically migrating files.

[0054] Please refer to Figure 1 , Figure 1 A flowchart of a file migration method disclosed by an embodiment of the present application is shown in the figure, and the method includes:

[0055] 101, obtain the current date and the client disk usage space corresponding to the current date.

[0056] In the embodiments of the present application, when the automatic migration file is performed, the current date and the client disk usage space corresponding to the current date can be obtained. The method for obtaining the client disk usage space corresponding to the current date can be scanning the client disk to obtain the basic information of the client disk, such as the total space of the client disk and the remaining space of the client disk, and then obtaining the client disk usage space by calculation, or obtaining the client disk usage space by other reasonable methods, which is not limited here.

[0057] 102. Inputting the current date into the pre-trained client disk usage rate regression model to obtain the client disk usage rate corresponding to the current date.

[0058] The current date can be input into the pre-trained client disk usage rate regression model to obtain the client disk usage rate corresponding to the current date.

[0059] 103. Inputting the current date into the pre-trained cloud disk usage rate regression model to obtain the cloud disk usage rate corresponding to the current date.

[0060] The current date can be input into the pre-trained cloud disk usage rate regression model to obtain the cloud disk usage rate corresponding to the current date.

[0061] 104. Determining whether the client disk usage space, the client disk usage rate corresponding to the current date, and the cloud disk usage rate corresponding to the current date meet the preset condition.

[0062] After obtaining the client disk usage space corresponding to the current date, the client disk usage rate corresponding to the current date, and the cloud disk usage rate corresponding to the current date, it can be determined whether the client disk usage space, the client disk usage rate corresponding to the current date, and the cloud disk usage rate corresponding to the current date meet the preset condition, and a determination result is obtained. If the preset condition is met, step 105 is performed, and if the preset condition is not met, subsequent steps are not performed. The preset condition can be that the client disk usage space is large, the client disk usage rate is large, and the cloud disk usage rate is small, and can also be any reasonable preset condition, which is not limited here.

[0063] 105. Obtaining the basic information of each to-be-migrated file in the client disk of the current date and the basic operation behavior information of the user.

[0064] The basic information of each to-be-migrated file in the client disk on the current day and the basic operation behavior information of the user can be obtained. Specifically, the client disk can be scanned to obtain the basic information, and other reasonable methods can also be used to obtain the basic information, and the specific method is not limited herein. It can be understood that the basic information of each to-be-migrated file can include file size, file creation time or file modification time, file opening times, and other basic information that can represent whether the user frequently uses the file, and the specific basic information is not limited herein. Secondly, the basic operation behavior information of the user can include the user's application usage list and the user's application usage duration for each application, and other information that can represent which applications the user does not frequently use, and the specific basic operation behavior information is not limited herein. Furthermore, the type of the to-be-migrated file can be an application data in addition to pictures or videos, and the basic operation behavior information of the user can determine the application file that the user does not frequently use, and the basic information of the to-be-migrated file such as the file creation time or the file modification time and the file opening times can also clearly determine the file that the user does not frequently use. Therefore, obtaining the basic information of each to-be-migrated file in the client disk on the current day and the basic operation behavior information of the user has certain significance for file migration.

[0065] 106. Determining the target migration file in the client disk according to the basic information of each to-be-migrated file and the basic operation behavior information of the user.

[0066] After obtaining the basic information of each to-be-migrated file in the client disk on the current day and the basic operation behavior information of the user, the target migration file in the client disk can be determined according to the basic information of each to-be-migrated file and the basic operation behavior information of the user. It can be understood that the shorter the duration of the user using an application, the more remote the file creation time or the file modification time of the application file, and the fewer the file opening times, which can indicate that the user does not frequently use the application file. Therefore, the method of determining the target migration file can be directly determining the target migration file according to the basic information of each to-be-migrated file and the basic operation behavior information of the user, or inputting the basic information of each to-be-migrated file and the basic operation behavior information of the user into a pre-trained logistic regression classification model to obtain the probability of not using each to-be-migrated file, or other reasonable methods of determining the target migration file, and the specific method is not limited herein.

[0067] 107. Migrating the target migration file to the cloud disk.

[0068] After determining the target migration file, the target migration file can be migrated to the cloud disk.

[0069] In the embodiments of the present application, it can be determined in real time whether the client meets the preset conditions for file migration. The target file to be migrated can be determined as long as the preset conditions are met, and the target file to be migrated is migrated to the cloud disk. The entire file migration process is automatically executed by the client, and the user does not perceive it. It can be ensured that the files in the client disk will not be too many, and there will be enough remaining available storage space in the client. The situation that the storage space of the client is not enough is reduced, and the user experience of using the client is improved.

[0070] In the embodiments of the present application, there are various methods for determining the target migration file in the client disk according to the basic information of each file to be migrated and the basic operation behavior information of the user. Based on the file migration method shown in Figure 1 One of the methods is described below.

[0071] In the embodiments, the date of the day and the client disk usage space corresponding to the date of the day can be obtained. Specifically, the date of the day, the total space of the client disk corresponding to the date of the day, and the remaining space of the client can be obtained. The difference between the total space of the client disk and the remaining space of the client is used as the client disk usage space. Other methods for obtaining the date of the day and the client disk usage space corresponding to the date of the day can also be used, which are not limited here.

[0072] The client disk usage rate regression model can be trained in advance. The specific training method can be as follows: first, obtain date samples of a preset time period, and the date samples are labeled with the client disk usage rate corresponding to the date samples. The client disk usage rate is determined by taking the ratio of the client disk usage space to the total space of the client disk as the client disk usage rate. Then, the date samples are input into the client disk usage rate regression model to obtain the predicted client disk usage rate output by the client disk usage rate regression model. Finally, the client disk usage rate loss between the predicted client disk usage rate and the labeled client disk usage rate is calculated according to the client disk usage rate regression loss function. When the client disk usage rate loss meets the first preset convergence condition, the training of the client disk usage rate regression model is completed. Other reasonable methods for training the client disk usage rate regression model can also be used, which are not limited here.

[0073] The cloud disk usage rate regression model can be pre-trained. A specific training method is as follows: first, a date sample of a preset time period is obtained, and the date sample is labeled with a cloud disk usage rate corresponding to the date sample, wherein the cloud disk usage rate is determined by taking a ratio of a cloud disk usage space to a total cloud disk space as the client disk usage rate; then, the date sample is input into the cloud disk usage rate regression model to obtain a predicted cloud disk usage rate output by the cloud disk usage rate regression model; finally, a cloud disk usage rate loss between the predicted cloud disk usage rate and the labeled cloud disk usage rate is calculated according to a cloud disk usage rate regression loss function. When the cloud disk usage rate loss satisfies a second preset convergence condition, a trained cloud disk usage rate regression model is obtained. Other reasonable training methods of the cloud disk usage rate regression model can also be used, and the specific training method is not limited herein.

[0074] After obtaining the trained client disk usage rate regression model and the trained cloud disk usage rate regression model, a current date can be input into the pre-trained client disk usage rate regression model to obtain a client disk usage rate corresponding to the current date, and the current date can be input into the pre-trained cloud disk usage rate regression model to obtain a cloud disk usage rate corresponding to the current date.

[0075] After obtaining the client disk usage space corresponding to the current date, the client disk usage rate corresponding to the current date, and the cloud disk usage rate corresponding to the current date, it can be determined whether the client disk usage space corresponding to the current date, the client disk usage rate corresponding to the current date, and the cloud disk usage rate corresponding to the current date satisfy the preset condition to obtain a determination result. The preset condition can be that the client disk usage space is greater than or equal to a preset usage space threshold, the client disk usage rate is greater than or equal to a preset client disk usage rate threshold, and the cloud disk usage rate is less than or equal to a preset cloud disk usage rate threshold. It can also be other reasonable preset conditions, which are not limited here. If the client disk usage space, the client disk usage rate corresponding to the current date, and the cloud disk usage rate corresponding to the current date satisfy the preset condition, the files to be migrated in the client disk can be determined. Specifically, the client disk can be scanned to obtain information about whether each file in the client disk is migrated. The files to be migrated in the client disk are determined according to the information about whether each file in the client disk is migrated. It can also be other methods for determining the files to be migrated in the client disk, which are not limited here. The files to be migrated can be files that need to read thumbnails, such as pictures, videos, PPTs, PDFs, and the like. They can also be application data. It should be understood that for the migration of pictures, the client disk needs to retain a thumbnail to deal with the case of network disconnection. When the upper application calls the kernel, it can first determine whether there is a network. If there is no network, the thumbnail can be read from the client disk, and the cloud disk will not be accessed through the HTTP protocol. If there is a network, the cloud disk can be accessed through the HTTP protocol. For the migration of videos, a video can be composed of multiple pictures. After migration to the cloud disk, the cloud disk can convert the video into a streaming media video. When the client's album clicks to play the video, it can be played in the form of streaming media. After playing, it can be saved to the local. It can also not be converted into a streaming media video. When the client's album clicks to play the video, it needs to be downloaded from the cloud disk to the client disk and then played. For the migration of non-media files such as documents and compressed packages, after migrating the file data of non-media files such as documents and compressed packages, if the network is disconnected, the client can see the existence of the non-media files such as documents and compressed packages, but cannot read the non-media files such as documents and compressed packages. For the migration of APP, the corresponding file to be migrated can be the file data of APP, and the whole data of APP. The icon and APP package name of APP are saved in the client disk, so that when the user clicks the APP, the whole data of APP can be downloaded to the client disk, so that the APP runs smoothly.

[0076] After determining the files to be migrated in the client disk, the basic information of each file to be migrated and the basic operation behavior information of the user in the client disk on the current day can be obtained, and the target migration file in the client disk can be determined according to the basic information of each file to be migrated and the basic operation behavior information of the user. The method of obtaining the basic information of each file to be migrated and the basic operation behavior information of the user can be a method of scanning the client disk, and can also be other reasonable obtaining methods, which are not limited here. The basic information of each file to be migrated can include file storage path, file name, file size, file creation time or file modification time, and file opening times, and the basic operation behavior information of the user can include the user's application use list and the user's application use duration of each application. The method of determining the target migration file in the client disk according to the basic information of each file to be migrated and the basic operation behavior information of the user can be various, and one of the methods will be described below. Please refer to Figure 2 , Figure 2 The flowchart of a method for determining a target migration file disclosed by an embodiment of the present application is shown in the figure, and the method comprises:

[0077] 201, each file to be migrated is sorted in the order from the first to the last according to the file creation time or the file modification time, and the first preset number of files to be migrated sorted in the front is screened to obtain a first screening set of files to be migrated.

[0078] Each file to be migrated can be sorted in the order from the first to the last according to the file creation time or the file modification time, and the first preset number of files to be migrated sorted in the front is screened to obtain a first screening set of files to be migrated. It can be understood that the earlier the time, the less frequently the file to be migrated is used, and the file to be migrated with a higher degree of infrequent use is screened out as the first screening.

[0079] 202, the basic information of each file to be migrated in the first screening set of files to be migrated and the basic operation behavior information of the user are input into a pre-trained logistic regression classification model to obtain the probability of not being used of each file to be migrated in the first screening set of files to be migrated.

[0080] The logistic regression classification model can be pre-trained. Specifically, the training method can be as follows: first, obtain the basic information sample of each to-be-migrated file sample and the basic operation behavior information sample of the user on each date sample in a preset time period; the basic information sample of each to-be-migrated file includes the file storage path, the file name, the file size, the file creation time or the file modification time, and the file opening times, etc., and the basic operation behavior information of the user includes the application use list of the user and the application use duration of the user using each application, etc., the file opening times greater than 0 are negative labels, and the file opening times equal to 0 are positive labels; then, the probability of not being used of each to-be-migrated file sample is calculated according to the file opening times of each to-be-migrated file sample, and is marked in the basic information sample of each to-be-migrated file sample; then, the basic information sample of each to-be-migrated file sample and the basic operation behavior information sample of the user are input into the logistic regression classification model, to obtain the predicted probability of not being used of each to-be-migrated file sample output by the logistic regression classification model; finally, the probability loss between the predicted probability and the marked probability is calculated according to the probability regression loss function, and when the probability loss satisfies a third preset convergence condition, the trained logistic regression classification model is obtained. Other reasonable training methods of the logistic regression classification model can also be used, and the specific training method is not limited herein.

[0081] After obtaining the first to-be-migrated file screening set and the pre-trained logistic regression classification model, the basic information of each to-be-migrated file in the first to-be-migrated file screening set and the basic operation behavior information of the user can be input into the pre-trained logistic regression classification model, to obtain the probability of not being used of each to-be-migrated file in the first to-be-migrated file screening set.

[0082] 203、The to-be-migrated files in the first to-be-migrated file screening set are sorted in descending order of probability, and a second preset number of to-be-migrated files in the front are screened, to obtain a second to-be-migrated file screening set.

[0083] After obtaining the probability of not being used of each to-be-migrated file in the first to-be-migrated file screening set, the to-be-migrated files in the first to-be-migrated file screening set are sorted in descending order of probability, and a second preset number of to-be-migrated files in the front are screened, to obtain a second to-be-migrated file screening set. It can be understood that the greater the probability is, the less frequently the to-be-migrated file is used, and the to-be-migrated files with a higher degree of infrequent use are screened again as the second screening.

[0084] 204、The to-be-migrated files in the second to-be-migrated file screening set are sorted in descending order of file size, and a third preset number of to-be-migrated files are screened, to obtain a target to-be-migrated file.

[0085] After obtaining the second filtered set of files to be migrated, each file in the second filtered set of files to be migrated can be sorted in descending order of file size, and the third preset number of files to be migrated can be filtered to obtain the target file to be migrated. It can be understood that after the previous two filtrations, it can be basically determined that the file to be migrated is not frequently used, and each file to be migrated has a large file and a small file. In order to increase the available space of the client disk as much as possible, the large file to be migrated can be filtered out first as the third filtration.

[0086] After determining the target file to be migrated, the target file to be migrated can be migrated to the cloud disk. Specifically, after determining the target file to be migrated, the kernel of the client can obtain a migration command of the target file to be migrated, can access the storage space where the file data in the target file to be migrated is located in the client disk according to the migration command of the target file to be migrated, and can migrate the target file to be migrated from the client disk to the cloud disk, and write a migration mark of the target file to be migrated in the client disk. The migration command of the target file to be migrated can be generated according to the operation of the user, can also be automatically generated by the computer device itself, and can also be generated in other ways. The method of generating the migration command is not limited here. For example, after the selection module of the computer device selects the target file to be migrated, the kernel is sent the migration command of the target file to be migrated. The selection module can be a neural network module, etc. The selection module can regularly or irregularly filter the files in the client disk. The method of migrating the target file to be migrated from the client disk to the cloud disk can be to encapsulate the file data of the target file to be migrated into a network data packet, send the network data packet to the cloud disk, and release the storage space where the file data of the target file to be migrated is located in the client disk. The method of releasing can be to mark the storage space originally storing the file data of the target file to be migrated as available space. This migration method of encapsulating the file data of the target file to be migrated into a network data packet through the HTTP protocol and sending it to the cloud disk can reduce the switching frequency between the user space and the kernel space, thereby improving the access speed of the client to the cloud disk. The cloud disk can be Baidu cloud disk, OneDrive, etc. which is not limited here.

[0087] It can be understood that, through three screenings, the least frequently used files to be migrated can be determined in a case that the available space of the client disk is as large as possible, and the least frequently used files to be migrated are migrated to the cloud disk, and the client can automatically migrate the files in the client disk without the intervention of the user, so that the client disk has sufficient storage space, thereby the client has sufficient remaining available storage space, the case that the storage space of the client is insufficient is reduced to the greatest extent, and the experience of the user using the client is improved to the greatest extent.

[0088] For the convenience of understanding the present embodiment, an example will be used to describe the method of the present embodiment.

[0089] (1) The method for training the client disk usage rate regression model and the cloud disk usage rate regression model is as follows:

[0090] One client corresponds to one user, and the client can scan the client disk to obtain the basic information of the client, such as date, user identifier of the user, total space of the client disk, remaining space of the client disk, total space of the cloud disk, and remaining space of the cloud disk. The difference between the total space of the client disk and the remaining space of the client disk is taken as the used space of the client disk, and the ratio of the used space of the client disk to the total space of the client disk is taken as the usage rate of the client disk. Similarly, the difference between the total space of the cloud disk and the remaining space of the cloud disk is taken as the used space of the cloud disk, and the ratio of the used space of the cloud disk to the total space of the cloud disk is taken as the usage rate of the cloud disk. Assuming that the preset time period is 30 days, the client disk usage rate and the cloud disk usage rate of the user per day within 30 days can be obtained, and the client disk usage rate and the cloud disk usage rate within 30 days are subjected to data fitting regression analysis. Specifically, the dates in 30 days are taken as independent variables x, and the client disk usage rate is taken as dependent variable y, and sample data pairs (x1, y1), (x2, y2),..., (x30, y30) are constructed. Similarly, the dates in 30 days are taken as independent variables x, and the cloud disk usage rate is taken as dependent variable z, and sample data pairs (x1, z1), (x2, z2),..., (x30, z30) are constructed. Nonlinear function f(x) = An*x^n + A(n-1)*x^(n-1) +... + A1*x + A0 is used for data fitting, and the client disk usage rate regression model and the cloud disk usage rate regression model corresponding to the user are obtained.

[0091] After the client disk usage rate regression model and the cloud disk usage rate regression model corresponding to the user are constructed, the client disk usage rate regression model and the cloud disk usage rate regression model corresponding to the user are continuously trained using sample data of the next 30 days of the user. Specifically, the method for training the client disk usage rate regression model is as follows: obtain the next 30 days of the user, each date is labeled with the client disk usage rate corresponding to the date, input each date into the client disk usage rate regression model, and obtain the predicted client disk usage rate of each date. For each date, the client disk usage rate loss between the predicted client disk usage rate and the labeled client disk usage rate is calculated according to the client disk usage rate regression loss function. When the client disk usage rate loss of each date meets the first preset convergence condition, the trained client disk usage rate regression model is obtained. The method for training the cloud disk usage rate regression model is similar to the method for training the client disk usage rate regression model, which is not described here.

[0092] (2) The method for training the logistic regression classification model is as follows:

[0093] One client corresponds to one user, and the client can scan the client disk to obtain the basic information of each to-be-migrated file in the client disk and the basic operation behavior information of the user. The basic information of the to-be-migrated file includes, for example, date, user identifier, file storage path, file size, file creation time or file modification time, and file opening times. The basic operation behavior information of the user includes, for example, the user's application use list and the user's application use duration for each application. The information of the user, such as the file storage path, the file size, the file creation time or the file modification time, the application use list, and the user's application use duration for each application, is integrated together with (date, user identifier) as the key, and the integrated information is taken as the sample feature, and the file opening times are taken as the label. The file opening times greater than 0 are negative labels, and the file opening times equal to 0 are positive labels. Assuming that the preset time period is 30 days, all sample data of the user within 30 days are aggregated, and the logistic regression classification model of the user is constructed according to all sample data of the user within 30 days.

[0094] After the user's logistic regression classification model is constructed, the user's next 30 days of sample data are used to train the user's corresponding logistic regression classification model. Specifically, the basic information sample of each to-be-migrated file sample and the user's basic operation behavior information sample of each date in the next 30 days of the user are obtained. The basic information sample of each to-be-migrated file sample includes file storage path, file name, file size, file creation time or file modification time, and file opening times, etc. The user's basic operation behavior information includes the user's application usage list and the user's application usage time length of each application, etc. The file opening times greater than 0 are negative labels, and equal to 0 are positive labels. The probability of not being used of each to-be-migrated file sample is calculated according to the file opening times of each to-be-migrated file sample, and is marked in the basic information sample of each to-be-migrated file sample. The basic information sample of each to-be-migrated file sample and the user's basic operation behavior information sample are input into the logistic regression classification model to obtain the predicted probability of not being used of each to-be-migrated file sample output by the logistic regression classification model. For each date, the probability loss between the predicted probability and the marked probability is calculated according to the probability regression loss function. When the probability loss of each date meets the third preset convergence condition, the trained logistic regression classification model is obtained.

[0095] (3) The method for determining the target to-be-migrated file is as follows:

[0096] The client first scans the client disk to obtain the migration information of each file on the client disk, the basic information of each to-be-migrated file, and the basic operation behavior information of the user. The basic information of each to-be-migrated file includes file storage path, file name, file size, file creation time or file modification time, and file opening times, etc. The basic operation behavior information of the user includes the user's application usage list and the user's application usage time length of each application, etc. The to-be-migrated file in the client disk is determined according to the migration information of each file on the client disk.

[0097] Then, the to-be-migrated file is screened three times, and the screening steps are as follows.

[0098] A, first screening:

[0099] The to-be-migrated files are sorted in the order of file creation time or file modification time from early to late, and 100 to-be-migrated files in the front are screened out.

[0100] B, second screening:

[0101] Inputting the basic information of 100 to-be-migrated files and the basic operation behavior information of the user into the pre-trained logistic regression classification model to obtain the probability that the 100 to-be-migrated files are not used, and sorting the 100 to-be-migrated files in descending order of the probability to select 50 to-be-migrated files in the front of the order.

[0102] C, third screening:

[0103] The 50 to-be-migrated files are sorted in descending order of file size, and 20 to-be-migrated files are selected to obtain 20 target to-be-migrated files.

[0104] In this embodiment, it can be determined in real time whether the client satisfies the preset condition of file migration. As long as the preset condition is satisfied, the target to-be-migrated file that needs to be migrated can be determined, and the target to-be-migrated file is migrated to the cloud disk. The entire file migration process is automatically executed by the client, and the user does not perceive it. It can be ensured that the files on the client disk are not too many, and there is enough remaining available storage space on the client. The situation that the storage space of the client is not enough is reduced, and the user experience of using the client is improved. Secondly, the client disk usage rate regression model and the cloud disk usage rate regression model are used, so that the client disk usage rate and the cloud disk usage rate can be quickly obtained, and the accuracy of the obtained client disk usage rate and the cloud disk usage rate is high. The implementability of automatically migrating files is improved. Thirdly, the logistic regression classification model is used, and the obtained probability accuracy is high, which improves the implementability of automatically migrating files. Finally, according to the basic information of each to-be-migrated file and the basic operation behavior information of the user, the to-be-migrated file is screened three times from different dimensions such as the creation and / or modification time dimension, the file opening times dimension, the file size dimension, and the user's application usage list dimension, the application usage time dimension and other different dimensions. The number of dimensions considered is more and more comprehensive, so that the to-be-migrated file that is used the least frequently can be determined while ensuring that the available space of the client disk is as large as possible. Thus, the client has enough remaining available storage space, the situation that the storage space of the client is not enough is reduced to the greatest extent, and the user experience of using the client is improved to the greatest extent.

[0105] The file migration method in the embodiments of the present application is described above, and the client in the embodiments of the present application is described below. Please refer to Figure 3 An embodiment of the client in the embodiments of the present application includes:

[0106] The obtaining unit 301 is configured to obtain the current date and the client disk usage space corresponding to the current date.

[0107] The input unit 302 is configured to input the date of the day obtained by the obtaining unit 301 into a pre-trained client disk usage rate regression model to obtain a client disk usage rate corresponding to the date of the day.

[0108] The input unit 302 is configured to input the date of the day obtained by the obtaining unit 301 into a pre-trained cloud disk usage rate regression model to obtain a cloud disk usage rate corresponding to the date of the day.

[0109] The judgment unit 303 is configured to, if the client disk usage space obtained by the obtaining unit 301, the client disk usage rate corresponding to the date of the day obtained by the input unit 302, and the cloud disk usage rate corresponding to the date of the day obtained by the input unit 302 satisfy a preset condition, obtain basic information of each to-be-migrated file in the client disk and basic operation behavior information of a user on the date of the day.

[0110] The determination unit 304 is configured to determine a target to-be-migrated file in the client disk according to the basic information of each to-be-migrated file and the basic operation behavior information of the user obtained by the judgment unit 303.

[0111] The migration unit 305 is configured to migrate the target to-be-migrated file determined by the determination unit 304 to a cloud disk.

[0112] In the embodiment of the present application, it can be determined in real time whether the client satisfies the preset condition of file migration. As long as the preset condition is satisfied, the target to-be-migrated file that needs to be migrated can be determined, and the target to-be-migrated file is migrated to the cloud disk. The whole file migration process is automatically executed by the client, and the user does not perceive it. It can be ensured that the files in the client disk are not too many, and there is enough remaining available storage space in the client. The situation that the storage space of the client is not enough is reduced, and the user experience of using the client is improved.

[0113] The client in the embodiment of the present application is described in detail below. Please refer to Figure 4 Another embodiment of the client in the embodiment of the present application includes:

[0114] The obtaining unit 401 is configured to obtain a date of the day and a client disk usage space corresponding to the date of the day.

[0115] The input unit 402 is configured to input the date of the day obtained by the obtaining unit 401 into a pre-trained client disk usage rate regression model to obtain a client disk usage rate corresponding to the date of the day.

[0116] The input unit 402 is configured to input the date of the day obtained by the obtaining unit 401 into a pre-trained cloud disk usage rate regression model to obtain a cloud disk usage rate corresponding to the date of the day.

[0117] The judgment unit 403 is configured to obtain the basic information of each to-be-migrated file and the basic operation behavior information of the user in the client disk on the current date if the client disk usage space obtained by the obtaining unit 401, the client disk usage rate corresponding to the current date obtained by the input unit 402, and the cloud disk usage rate corresponding to the current date obtained by the input unit 402 meet a preset condition.

[0118] The determination unit 404 is configured to determine the target migration file in the client disk according to the basic information of each to-be-migrated file and the basic operation behavior information of the user obtained by the judgment unit 403.

[0119] The migration unit 405 is configured to migrate the target migration file determined by the determination unit 404 to the cloud disk.

[0120] The client further includes a calculation unit 406.

[0121] The obtaining unit 401 is further configured to obtain a date sample in a preset time period, and the date sample is labeled with a client disk usage rate corresponding to the date sample.

[0122] The input unit 402 is further configured to input the date sample into a client disk usage rate regression model to obtain a predicted client disk usage rate output by the client disk usage rate regression model.

[0123] The calculation unit 406 is configured to calculate a client disk usage rate loss between the predicted client disk usage rate and the labeled client disk usage rate according to a client disk usage rate regression loss function, and obtain a trained client disk usage rate regression model when the client disk usage rate loss meets a first preset convergence condition.

[0124] The obtaining unit 401 is further configured to obtain a date sample in a preset time period, and the date sample is labeled with a cloud disk usage rate corresponding to the date sample.

[0125] The input unit 402 is further configured to input the date sample into a cloud disk usage rate regression model to obtain a predicted cloud disk usage rate output by the cloud disk usage rate regression model.

[0126] The calculation unit 406 is further configured to calculate a cloud disk usage rate loss between the predicted cloud disk usage rate and the labeled cloud disk usage rate according to a cloud disk usage rate regression loss function, and obtain a trained cloud disk usage rate regression model when the cloud disk usage rate loss meets a second preset convergence condition.

[0127] The client further includes:

[0128] The scanning unit 407 is configured to scan the client disk to obtain information about whether each file of the client disk is to be migrated.

[0129] The determining unit 404 is further configured to determine the files to be migrated in the client disk according to the information about whether each file of the client disk is to be migrated scanned by the scanning unit 407.

[0130] The determining unit 404 is specifically configured to sort the each file to be migrated in the order from the file creation time or the file modification time to the front, filter a first preset number of files to be migrated in the front of the order, obtain a first filtered set of files to be migrated, input the basic information of each file to be migrated in the first filtered set of files to be migrated and the basic operation behavior information of the user into a pre-trained logistic regression classification model, obtain the probability that each file to be migrated in the first filtered set of files to be migrated is not used, sort each file to be migrated in the first filtered set of files to be migrated in the order from the probability from large to small, filter a second preset number of files to be migrated in the front of the order, obtain a second filtered set of files to be migrated, sort each file to be migrated in the second filtered set of files to be migrated in the order from the file size from large to small, filter a third preset number of files to be migrated, and obtain the target file to be migrated.

[0131] The obtaining unit 401 is further configured to obtain the basic information sample of each file to be migrated sample and the basic operation behavior information sample of the user of each date sample in a preset time period; the basic information sample of each file to be migrated sample includes a file saving path, a file name, a file size, a file creation time or a file modification time, and a file opening number, etc., and the basic operation behavior information of the user includes an application use list of the user and an application use duration of the user using each application, etc., the file opening number greater than 0 is a negative label, and the file opening number equal to 0 is a positive label.

[0132] The calculating unit 402 is further configured to calculate the probability that each file to be migrated sample is not used according to the file opening number of each file to be migrated sample, and mark the probability in the basic information sample of each file to be migrated sample.

[0133] The input unit 402 is further configured to input the basic information sample of each file to be migrated sample and the basic operation behavior information sample of the user into a logistic regression classification model to obtain a predicted probability that each file to be migrated sample is not used output by the logistic regression classification model.

[0134] The computing unit 406 is further configured to calculate a probability loss between the predicted probability and the labeled probability according to a probability regression loss function, and obtain a trained logistic regression classification model when the probability loss satisfies a third preset convergence condition.

[0135] In the embodiment, each unit in the client performs the operation of the client in the embodiment as shown in Figure 1 and Figure 2 will not be described here in detail.

[0136] Please refer to Figure 5 , another embodiment of the client 500 in the embodiment of the application includes:

[0137] The central processing unit 501, the memory 505, the input and output interface 504, the wired or wireless network interface 503 and the power supply 502;

[0138] The memory 505 is a volatile storage memory or a persistent storage memory;

[0139] The central processing unit 501 is configured to communicate with the memory 505 and execute the instruction operation in the memory 505 to perform the method in the embodiment as shown in Figure 1 and Figure 2 .

[0140] The embodiment of the application further provides a computer readable storage medium, which includes instructions, when the instructions run on a computer, make the computer execute the method in the embodiment as shown in Figure 1 and Figure 2 .

[0141] The embodiment of the application further provides a computer program product including instructions, when the computer program product runs on a computer, make the computer execute the method in the embodiment as shown in Figure 1 and Figure 2 .

[0142] It should be understood that, although each step in the flowchart involved in each embodiment as described above is displayed in sequence according to the arrow, these steps are not necessarily executed in sequence according to the arrow. Unless explicitly stated in this article, the execution of these steps has no strict sequence limitation, and these steps can be executed in other orders. Moreover, at least part of the steps in the flowchart involved in each embodiment as described above can include multiple steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence of these steps or stages is not necessarily sequential, but can be executed in rotation or alternation with at least part of other steps or steps or stages in other steps.

[0143] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described system, device and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be repeated here.

[0144] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other ways. For example, the above-described device embodiments are only schematic, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or other forms.

[0145] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0146] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0147] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, read-only memory), a random access memory (RAM, random access memory), a magnetic disk or an optical disk, and various media that can store program codes.

Claims

1. A file migration method characterized by, The method is applied to a client and comprises the following steps: obtaining a current date and a client disk usage space corresponding to the current date; inputting the current date into a pre-trained client disk usage rate regression model to obtain a client disk usage rate corresponding to the current date, wherein the client disk usage rate represents a degree of occupation of a local disk space of the client; inputting the current date into a pre-trained cloud disk usage rate regression model to obtain a cloud disk usage rate corresponding to the current date, wherein the cloud disk usage rate represents a degree of occupation of a cloud storage space; if the client disk usage space, the client disk usage rate corresponding to the current date and the cloud disk usage rate corresponding to the current date satisfy a preset condition, obtaining basic information of each to-be-migrated file in the client disk and basic operation behavior information of a user on the current date, wherein the preset condition is that the client disk usage space is greater than or equal to a preset usage space threshold, the client disk usage rate is greater than or equal to a preset client disk usage rate threshold, and the cloud disk usage rate is less than or equal to a preset cloud disk usage rate threshold; the basic information of each to-be-migrated file comprises a file storage path, a file name, a file size, a file creation time or a file modification time and a file opening frequency, and the basic operation behavior information of the user comprises an application use list of the user and an application use duration of the user for each application; determining a target migration file in the client disk according to the basic information of each to-be-migrated file and the basic operation behavior information of the user; and migrating the target migration file to a cloud disk.

2. The file migration method of claim 1, wherein, Before the step of inputting the current date into the pre-trained client disk usage rate regression model to obtain the client disk usage rate corresponding to the current date, the method further comprises the following steps: obtaining date samples in a preset time period, wherein the date samples are labeled with client disk usage rates corresponding to the date samples; inputting the date samples into a client disk usage rate regression model to obtain predicted client disk usage rates output by the client disk usage rate regression model; calculating a client disk usage rate loss between the predicted client disk usage rates and the labeled client disk usage rates according to a client disk usage rate regression loss function, and obtaining a trained client disk usage rate regression model when the client disk usage rate loss satisfies a first preset convergence condition.

3. The file migration method of claim 1, wherein, Before the step of inputting the current date into the pre-trained cloud disk usage rate regression model to obtain the cloud disk usage rate corresponding to the current date, the method further comprises the following steps: obtaining date samples in a preset time period, wherein the date samples are labeled with cloud disk usage rates corresponding to the date samples; inputting the date samples into a cloud disk usage rate regression model to obtain predicted cloud disk usage rates output by the cloud disk usage rate regression model; calculating a cloud disk usage rate loss between the predicted cloud disk usage rates and the labeled cloud disk usage rates according to a cloud disk usage rate regression loss function, and obtaining a trained cloud disk usage rate regression model when the cloud disk usage rate loss satisfies a second preset convergence condition.

4. The file migration method of claim 1, wherein, Before the method further includes obtaining the basic information of each to-be-migrated file in the client disk on the current date and the basic operation behavior information of the user, the method further includes: Scanning the client disk to obtain information about whether each file in the client disk is to be migrated; According to the information about whether each file in the client disk is to be migrated, determining the to-be-migrated file in the client disk.

5. The file migration method according to any one of claims 1 to 4, characterized by, The method further includes: Sorting the to-be-migrated files according to the file creation time or the file modification time from early to late, and screening the first preset number of to-be-migrated files in the front to obtain a first screened set of to-be-migrated files; Inputting the basic information of each to-be-migrated file in the first screened set of to-be-migrated files and the basic operation behavior information of the user into a pre-trained logistic regression classification model to obtain the probability that each to-be-migrated file in the first screened set of to-be-migrated files is not used; Sorting each to-be-migrated file in the first screened set of to-be-migrated files according to the probability from large to small, and screening the second preset number of to-be-migrated files in the front to obtain a second screened set of to-be-migrated files; Sorting each to-be-migrated file in the second screened set of to-be-migrated files according to the file size from large to small, and screening the third preset number of to-be-migrated files to obtain the target migration file.

6. The file migration method of claim 5, wherein, Before the method further includes inputting the basic information of each to-be-migrated file in the first screened set of to-be-migrated files and the basic operation behavior information of the user into a pre-trained logistic regression classification model, the method further includes: Obtaining the basic information sample of each to-be-migrated file sample and the basic operation behavior information sample of the user for each date sample in a preset time period; the basic information sample of each to-be-migrated file includes the file storage path, the file name, the file size, the file creation time or the file modification time, and the file opening times, and the basic operation behavior information of the user includes the application use list of the user and the application use duration of the user using each application, and the file opening times greater than 0 are negative labels and equal to 0 are positive labels; According to the file opening times of each to-be-migrated file sample, calculating the probability that each to-be-migrated file sample is not used, and labeling the basic information sample of each to-be-migrated file sample; Inputting the basic information sample of each to-be-migrated file sample and the basic operation behavior information sample of the user into a logistic regression classification model to obtain the predicted probability of each to-be-migrated file sample not being used output by the logistic regression classification model; According to the probability loss function, calculating the probability loss between the predicted probability and the labeled probability, and when the probability loss satisfies a third preset convergence condition, obtaining a trained logistic regression classification model.

7. A client, characterized in that, The method further includes: An obtaining unit is configured to obtain the client disk usage space corresponding to the current date; An input unit is configured to input the current date into a pre-trained client disk usage rate regression model to obtain the client disk usage rate corresponding to the current date; The client disk usage rate represents a degree to which a local disk space of the client has been occupied. The input unit is further configured to input the current date into the pre-trained cloud disk usage rate regression model to obtain a cloud disk usage rate corresponding to the current date. The cloud disk usage rate represents a degree to which a cloud storage space has been occupied. The judgment unit is configured to obtain, if the client disk usage space, the client disk usage rate corresponding to the current date, and the cloud disk usage rate corresponding to the current date satisfy a preset condition, basic information of each to-be-migrated file in the client disk and basic operation behavior information of a user in the current date, the preset condition being that the client disk usage space is greater than or equal to a preset usage space threshold, the client disk usage rate is greater than or equal to a preset client disk usage rate threshold, and the cloud disk usage rate is less than or equal to a preset cloud disk usage rate threshold, the basic information of each to-be-migrated file including a file storage path, a file name, a file size, a file creation time or a file modification time, and a file opening number, and the basic operation behavior information of the user including an application use list of the user and an application use duration of the user using each application. The determination unit is configured to determine a target migration file in the client disk according to the basic information of each to-be-migrated file and the basic operation behavior information of the user. The migration unit is configured to migrate the target migration file to a cloud disk.

8. A client, characterized in that, The computer readable storage medium comprises instructions which, when executed on a computer, cause the computer to perform the method of any one of claims 1 to 6. The computer readable storage medium comprises instructions which, when executed on a computer, cause the computer to perform the method of any one of claims 1 to 6. ​ ​ 9. A computer-readable storage medium, characterized in that, ​

Images