WO2019228134A1 - Application program pre-loading method and device, storage medium, and terminal - Google Patents

Abstract

Disclosed is an application program pre-loading method, comprising: when it is detected that an application pre-loading event is triggered, determining a target candidate application set by using a first type of prediction model; determining a target application program to be pre-loaded comprised in the target candidate application set by using a second type of prediction model corresponding to the target candidate application set; and pre-loading the target application program. Also disclosed are an application program pre-loading device, a storage medium, and a terminal.

Description

Application program preloading method, device, storage medium and terminal

This application claims priority from a Chinese patent application filed with the Chinese Patent Office on May 29, 2018 with application number 201810532724.9, the entire contents of which are incorporated herein by reference.

Technical field

The embodiments of the present application relate to the technical field of application program loading, for example, to a method, a device, a storage medium, and a terminal for preloading an application program.

Background technique

At present, terminals such as smart phones, tablet computers, laptops, and smart home appliances have become essential electronic devices in people's daily lives. With the continuous intelligentization of terminal devices, most terminal devices are loaded with operating systems, enabling terminal devices to install a variety of applications to meet different needs of users.

With the continuous improvement of the configuration of terminal devices, dozens or even hundreds of applications can be installed in most terminal devices, and as the functions of applications become more and more abundant, more resources need to be loaded when the applications run. When the user chooses to start an application, the terminal will load the resources required for the application to start. After loading, the initial interface of the application is entered. The entire process usually takes several seconds or even ten seconds, resulting in the application. The start-up efficiency of the program is low and urgently needs improvement.

Summary of the invention

The embodiments of the present application provide a method, a device, a storage medium, and a terminal for preloading an application program, which can optimize a preloading solution for the application program.

In an embodiment, an embodiment of the present application provides a method for preloading an application program, including:

When it is detected that the application preload event is triggered, the first type of prediction model is used to determine the target candidate application set;

Determining a target application to be preloaded included in the target candidate application set by using a second type prediction model corresponding to the target candidate application set;

Pre-load the target application.

In an embodiment, an embodiment of the present application provides an application preloading device, including:

The application set determination module is configured to determine a target candidate application set by using a first-type prediction model when an application preload event is detected to be triggered;

A target application determination module configured to determine a target application to be preloaded in the target candidate application set by using a second type prediction model corresponding to the target candidate application set;

The preloading module is configured to preload the target application.

In an embodiment, a computer-readable storage medium is provided in an embodiment of the present application. The storage medium stores a computer program, and when the computer program is executed by a processor, the application program according to the embodiment of the present application is preloaded. method.

In an embodiment, an embodiment of the present application provides a terminal, including a memory, a processor, and a computer program stored in the memory and executable by the processor. When the processor executes the computer program, the implementation is as in the present application. The application preloading method described in the embodiment.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic flowchart of an application preloading method according to an embodiment of the present application; FIG.

2 is a schematic diagram of a relative position relationship between a preloaded active window stack and a display screen display area according to an embodiment of the present application;

3 is a schematic diagram of a relative position relationship between another preloaded active window stack and a display screen display area according to an embodiment of the present application;

4 is a schematic diagram of application interface migration provided by an embodiment of the present application;

5 is a schematic flowchart of another application preloading method according to an embodiment of the present application;

6 is a schematic flowchart of another application preloading method according to an embodiment of the present application;

FIG. 7 is a structural block diagram of an application preloading device according to an embodiment of the present application; FIG.

8 is a schematic structural diagram of a terminal according to an embodiment of the present application;

FIG. 9 is a schematic structural diagram of another terminal according to an embodiment of the present application.

Detailed ways

The technical solution of the present application will be described below with reference to the drawings and embodiments. It can be understood that the embodiments described herein are only used to explain the present application, rather than limiting the present application. For ease of description, only some, but not all, structures related to the present application are shown in the drawings.

Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although flowcharts describe multiple steps as sequential processing, many of these steps can be performed in parallel, concurrently, or simultaneously. In addition, the order of one or more steps may be rearranged. The process may be terminated when its operation is completed, but may also have additional steps not included in the drawings. The processing may correspond to methods, functions, procedures, subroutines, subroutines, and so on.

FIG. 1 is a schematic flowchart of an application preloading method according to an embodiment of the present application. The method may be executed by an application preloading device, where the device may be implemented by software and / or hardware, and may generally be integrated in a terminal. As shown in Figure 1, the method includes:

Step 110: When it is detected that an application preloading event is triggered, a first type prediction model is used to determine a target candidate application set.

Exemplarily, the terminal in the embodiments of the present application may include terminal devices such as a mobile phone, a tablet computer, a notebook computer, and smart home appliances. The terminal is loaded with an operating system.

Exemplarily, the triggering condition of the application preload event can be set according to the actual situation, which is not specifically limited in the embodiment of the present application. For example, an application preload event can be triggered when it is detected that a user's action meets a preset condition (such as picking up a terminal, entering a screen to unlock an operation, or inputting a terminal to unlock an operation, etc.); or when a change to a foreground application is detected, Trigger an application preload event; or you can trigger an application preload event immediately (or after a preset time period has elapsed) after the prediction process of the preloaded application ends; or you can trigger at regular intervals. After the application preload event is triggered, the system can detect that the application preload event has been triggered by reading a flag bit or receiving a trigger instruction, and the specific detection method is not limited in this embodiment of the present application.

In the embodiment of the present application, before detecting that an application preload event is triggered, or before detecting whether an application preload event is triggered, a candidate application installed in a terminal may be divided to obtain multiple candidate application sets. Candidate applications can include all applications installed in the terminal, or they can include some applications. Some applications may include applications that are frequently used by users, and may also include third-party applications, that is, some applications may not include applications that are rarely used by users, or may not include system applications. This embodiment of the present application does not limit the manner and quantity of candidate application programs. In an embodiment, the candidate application can be determined according to the number of times and / or the use time of one or more applications in a preset period before the current time. The preset period is, for example, 1 month, when the number of uses and / or When the usage time exceeds the corresponding threshold, the corresponding application is determined as a candidate application, or multiple applications are sorted according to the number of uses and / or the length of use, and the top-ranked application is determined as a candidate application .

In the embodiment of the present application, the division manner of the candidate application set is not limited. For example, it can be divided according to application types, that is, applications that belong to the same type are divided into candidate application sets corresponding to the type, and the application rules for dividing application types are not limited, for example, according to the user ’s Personal needs or the default classification in the app store are divided into social, office, game, shopping, property, photography and video, and education; etc .; also according to the frequency of application use during the historical use period For example, if multiple use frequency intervals are set, they are classified into a certain frequency interval according to the application frequency of the application in the historical period (for example, in the past month), and each frequency interval corresponds to a candidate application set; The training samples used for model training are clustered. For example, K-NearestNeighbor (KNN) algorithm is used for clustering, and the candidate application set is divided according to the clustering result. It can also be divided according to folders, that is, belong to the same An application in a folder is divided into a set of candidate applications corresponding to the folder. The name is the corresponding candidate application set; it can also be divided by the desktop interface, that is, the application corresponding to the application icon belonging to the same desktop interface is divided into the candidate application set corresponding to the desktop interface, which can be based on the desktop. The sequence number names the corresponding candidate application set. In addition, the division method of the candidate application set can also be freely set by the user. For example, a division list is maintained in the terminal, and the user can add applications for each candidate application set according to his actual needs.

In the embodiment of the present application, the first type of prediction model is set to predict which candidate application set the application program to be started by the user belongs to. The first type of prediction model can be a machine learning model. The algorithms used can include Recurrent Neural Networks (RNN), Long Short-Term Memory (LSTM) networks, and Deep Neural Networks. (DNN), threshold loop unit, simple loop unit, autoencoder, decision tree, random forest, feature mean classification, classification regression tree, hidden Markov, K-Nearest Neighbor (KNN) algorithm, logistic regression model, Bayesian model, Gaussian model, KLback-Leibler divergence, etc.

In an embodiment, during the process of using the terminal by the user, a first training sample may be collected, and a sample mark corresponding to the first training sample may be recorded. The sample mark is a candidate application set to which the application program opened at the time of collecting the training sample, Or after the sample collection time (can be within a set time period, such as within 10 minutes), the candidate application set to which the opened application belongs, uses the first training sample and the corresponding sample mark to train the preset initial model, and finally gets used for The first type of prediction model that predicts which candidate application set the application to be preloaded belongs to. Exemplarily, the elements included in the first training sample may include the time, place, and frequency of the application being opened; and may include the operating status of the terminal, such as the on / off status of the mobile data network, the connection status of the wireless hotspot, and the connected Wireless hotspot identity information, currently running application, previous foreground application, the length of time the current application stays in the background, the time when the current application was last switched to the background, the plugging state of the headphone jack, the charging status, Battery power information and screen display time; it can also include data collected by sensors integrated in the terminal, such as motion sensors, light sensors, temperature sensors, and humidity sensors.

For example, the appropriate sample elements can be selected according to the selected machine learning model, and the selected machine learning model can be determined according to the selected sample elements. The model can also be combined with the needs for prediction accuracy and prediction speed. And the selection of sample elements are not limited in the embodiments of the present application.

Exemplarily, the first current sample may be collected according to the sample elements included in the first training sample, the first current sample is input into the first type prediction model, and the target candidate application is determined according to the output result of the first type prediction model. set. In an embodiment, the output result of the first type of prediction model may be the hit probability of each candidate application set, and the candidate application set with the higher or highest hit probability is determined as the target candidate application set. The target candidate application set may be one or more, which is not limited in the embodiment of the present application.

Step 120: Use a second-type prediction model corresponding to the target candidate application set to determine a target application to be preloaded included in the target candidate application set.

In the embodiment of the present application, before detecting that an application preload event is triggered, or before detecting whether an application preload event is triggered, model training is performed for each candidate application set to obtain a second corresponding to each candidate application set. A class prediction model, and the second type of prediction model is set to predict an application program to be started by a user included in a corresponding candidate application set (that is, a target candidate application set). The second type of prediction model can be a machine learning model. The algorithms used can include Recurrent Neural Networks (RNN), Deep Neural Network (DNN), Long Short-Term Memory (LSTM) ) Network, threshold loop unit, simple loop unit, autoencoder, decision tree, random forest, feature mean classification, classification regression tree, hidden Markov, K-Nearest Neighbor (KNN) algorithm, logistic regression model, Bayesian model, Gaussian model, KLback-Leibler divergence, etc. The algorithm used for the second type of prediction model may be the same as or different from the first type of prediction model, which is not limited in the embodiment of the present application.

In an embodiment, during the process of using the terminal by the user, a second training sample may be collected and a sample mark corresponding to the second training sample may be recorded. The sample mark is an application program opened at the time of training sample collection, or after the time of sample collection (Can be within a set period of time, such as within 10 minutes) for an open application (the opened application belongs to the corresponding candidate application set), the first training sample and the corresponding sample mark are used to predict the current candidate application set corresponding to the current candidate application set. The initial model is set for training, and finally a second type of prediction model for predicting the application to be preloaded is obtained. Exemplarily, the elements included in the second training sample may include the time, place, and frequency of the application being opened; and may include the operating status of the terminal, the on / off status of the mobile data network, the connection status of the wireless hotspot, and the connected wireless Hotspot identity information, currently running application, previous foreground application, length of time the current application stays in the background, time when the current application was last switched to the background, plug and unplug status of the headphone jack, charging status, battery Power information and screen display time; it can also include data collected by sensors integrated in the terminal, such as motion sensors, light sensors, temperature sensors, and humidity sensors. In the embodiment of the present application, the elements included in the second training sample may be the same as or different from the elements included in the first training sample. In some embodiments, the number of element types included in the first training sample is less than the number of element types included in the second training sample. The advantage of this setting is that because the first type prediction model is used for rough prediction, the number of sample bodies can be reduced. To reduce the time cost of model training and applying the model to make predictions, and improve the prediction speed.

Exemplarily, the second current sample may be collected according to the sample elements included in the second training sample, and the second current sample may be input into a second type prediction model corresponding to the target candidate application set. The output determines the target application to be preloaded. In an embodiment, the output result of the second type of prediction model may be the startup probability of each application corresponding to the candidate application set (ie, the target candidate application set), and the application with the higher or highest startup probability is determined as the target application. program. The startup probability includes a probability that an application program is about to be opened.

Step 130: Preload the target application program.

In the embodiment of the present application, when there is only one target application, there is no need to consider the preloading order of the applications; and when there are multiple target applications, multiple target applications may be determined one by one as the current to-be-loaded application, Pre-loading operations are performed in sequence, or more than two target applications can be determined as the current to-be-pre-loaded applications, and pre-loading operations are performed simultaneously, that is, the pre-loading process of multiple applications can be performed in parallel.

In the embodiment of the present application, the specific process of preloading and the resources loaded are not limited. For example, a corresponding hardware resource may be allocated to a target application to be preloaded, and related data required for startup may be loaded based on the allocated hardware resources. Exemplarily, it may include application process startup, application service startup, memory allocation, file content reading, network data acquisition, and interface rendering. In addition, pre-loaded resources can be determined according to the specific type of application to be pre-loaded. For example, if the application to be preloaded is a social software, you can preload the startup screen, contact list, and recent message records in the application; if the application to be preloaded is a game, you can preload the application Game background-related data in the app.

In one embodiment, after the target application is preloaded, when receiving a running instruction of the target application, the target application corresponding to the running instruction is started based on the preloaded resource.

In the application preloading method provided in the embodiment of the present application, when it is detected that an application preloading event is triggered, the first type of prediction model is used to determine the target candidate application set, and then the second type of prediction model corresponding to the target candidate application set is used to determine the target. The target application to be preloaded included in the candidate application set is used to preload the target application. By adopting the above technical solution, two different types of prediction models can be used for staged prediction, which improves the accuracy of application prediction.

In some embodiments, the determining a target candidate application set by using the first type of prediction model includes: obtaining a current use time-series association sequence of a foreground application, the current use time-series association sequence including the foreground application and the A sequence formed by at least one application previously used by the foreground application in chronological order; inputting the currently used time-series correlation sequence into a first-type prediction model, and determining a target candidate application according to an output result of the first-type prediction model set. Exemplarily, when detecting that the application B is switched to the application A, it indicates that the switching operation of the application running in the foreground is detected. At this time, the first application currently running is the application A, and the application running at the last moment is Application B, then the currently used time-series association sequence of the currently running first application A is: application B-application A. As another example, the first application currently running is application C, the application running at the last moment is application D, and the application running at the moment before the last moment is application E. It can be understood that, first, the application E is switched to the application D, and then the application D is switched to the application C, and then the currently used time sequence correlation sequence of the currently running application C can be expressed as: application E-application D-Application C. The embodiment of the present application does not limit the number of applications included in the current usage time-series association sequence of the foreground application. The advantage of this setting is that each time the current application changes, the pre-sequence usage sequence corresponding to the current foreground application changes, and based on the order in which the applications are opened, the application that the user will use quickly and accurately can be predicted The set of candidate applications to which the program belongs.

In some embodiments, before the detecting that an application preload event is triggered, the method further includes: collecting a historical use time-series correlation sequence of the sample application within a first preset time period, as the first training sample, the first A sample tag of a training sample includes a candidate application set to which an application program used after the sample application program belongs; and inputting the first training sample and a corresponding sample tag into a first preset model to the The first preset model is trained, and the trained model is used as the first type of prediction model. In an embodiment, the order in which multiple applications are opened in the terminal may be recorded in a preset history period, that is, when the foreground application is switched, the switched application is recorded to obtain an application sequence. The RNN network is used as an example for illustration, and the step size used is 2, so when the sample is extracted, every 2 applications can form a training sample, and the following application labels the sample of the training sample. For example, the stored application sequence is 1,2,2,5,3 ..., where the numbers represent the application's number, which can be disassembled into samples (1,2 → 2), (2,2 → 5), (2,5 → 3), ..., where the first two numbers in each sample are input, and the number after the → symbol is the corresponding sample label. The training samples are sequentially input into a preset RNN network to obtain a first-type prediction model.

In some embodiments, the determining a target application to be preloaded included in the target candidate application set by using a second type prediction model corresponding to the target candidate application set includes: obtaining current state characteristic information of the terminal; Inputting the current state characteristic information into a second type prediction model corresponding to the target candidate application set, and determining a target to be preloaded in the target candidate application set according to an output result of the second type prediction model application. The advantage of this setting is that after using the application sequence to predict the target candidate application set, a richer sample element can be used for further accurate prediction. The current state characteristic information may include the time, place, and frequency of the application being opened; it may include the operating state of the terminal, such as the on / off state of the mobile data network, the connection status of the wireless hotspot, the identity information of the connected wireless hotspot, The plugging and unplugging status, charging status, battery power information, and screen display duration of the headphone jack; it can also include data collected by sensors integrated in the terminal, such as motion sensors, light sensors, temperature sensors, and humidity sensors.

In some embodiments, before the detecting that an application preload event is triggered, the method further includes: collecting a historical state characteristic of the terminal when a sample application in a current candidate application set is used in a second preset time period. Information, as the second training sample, the sample tag of the second training sample includes the sample application or an application program used after the sample application; inputting the second training sample and a corresponding sample tag In the second preset model, the second preset model is trained, and the trained second preset model is used as a second type prediction model corresponding to the current candidate application set. The advantage of this setting is that the model is trained with richer sample elements, which improves prediction accuracy and accuracy.

In the embodiment of the present application, the training process of the first type prediction model and the second type prediction model may be performed locally on the terminal, or may be performed on other devices such as a server, which is not limited in the embodiment of the present application.

In an embodiment, the first preset time period and the second preset time period may be the same or different.

In some embodiments, the pre-loading the target application includes pre-loading an application interface corresponding to the target application based on a pre-created pre-loading active window stack, wherein the pre-loading active window stack The corresponding boundary coordinates are outside the coordinate range of the display screen. The advantage of this setting is that pre-loading the application interface can greatly complete the preparation work before the application is started, and improve the startup speed of the application to be pre-loaded without affecting the display content of the foreground application on the display screen. On the display.

In the embodiment of the present application, the active window may be understood as an independent interface that directly provides users with interaction and operation, and different terms may be used to name the interface in different operating systems. In order to facilitate understanding, the following description is made by taking an Android operating system as an example.

In the Android system, the active window is called Activity. Activity is a component responsible for interacting with the user. It provides a screen (which can be understood as a screen interface instead of a physical display screen) for the user to interactively complete a certain task. In an android application, an Activity is usually a separate screen, which can display some controls or listen to and handle user events. In the management of Activity, there are two concepts: Task stack (Task) and activity window stack (Stack). Task corresponds to an application. Task is set to store Activity. One Task can store one or more Activities, and these Activities follow the principle of "first in, first out, last in, first out". The Stack is set to manage Tasks. Generally, a Stack manages the Tasks belonging to one or more Activities that need to be displayed on a screen. A Stack can manage one or more Tasks. Of course, the Stack also follows the basics of the Stack. Management principles. The screens described here are not necessarily complete and independent display screens. Taking "two screens" as an example, these two screens may be just two areas in a complete display screen that independently display their respective display contents. Of course, if the terminal has two or more independent display screens, the "two screens" may also be two independent display screens.

In the Android system, multi-window mode is supported, which can include split-screen mode, picture-in-picture mode, and FreeForm. In the multi-window mode, the stack where the application is located can have its own size, which can include the upper, lower, left, and right coordinates in a coordinate system with the upper left corner of the terminal screen as the origin. For example, (a, b, c, d) generally describes a rectangular boundary, which can be expressed by the coordinates of the upper left corner and the lower right corner of the rectangle, that is, the upper left corner of the rectangle is (a, b), right The lower corner coordinate is (c, d), and such a rectangular area corresponds to the size of the Stack. The layout of the application in the stack is based on the size of the stack, which means that the application interface corresponding to the activity is displayed within the boundary of the size.

In multi-window mode, you can allow multiple applications to be visible, including both system and user and system only. Both visible to the system and the user means to be displayed on the display screen and can be seen by the user; only visible to the system means that the operating system is visible and the user is not visible, which may be obscured by the application in the foreground or as implemented in this application. Show off screen.

In the embodiment of the present application, pre-loading the application interface of the target application program outside the display screen can be implemented based on the multi-window mechanism of the operating system, and the size corresponding to the application program is set outside the display screen through the multi-window mechanism to achieve The purpose is not visible to the user, so that it does not affect the display of the foreground application display content on the display screen.

Generally, in multi-window mode, there can be multiple types of stacks. For example, Home Stack represents the stack displayed by desktop applications, App Stack represents the stack displayed by third-party applications, and there can be other split-screen stacks. The contained content can be displayed on a display screen, which is collectively referred to as an application active window stack in the embodiments of the present application. In the embodiment of the present application, a preloaded active window stack (preloaded stack) is added, which is set to represent the stack displayed by the preloaded application, and the boundary coordinates of the preloaded stack are set outside the coordinate range of the display screen, to be preloaded The application can be displayed on that Stack. For the Android system, based on the multi-window mechanism of the Android system, a new Stack dedicated to displaying preloaded applications can be created. In the embodiment of the present application, a new stack is created because the newly created preloaded stack can have its own size and visibility, thereby achieving the purpose of preloading outside the display screen.

In the embodiment of the present application, the creation timing of the preloaded stack is not limited. The preloaded stack is set to the resident state by default before the terminal leaves the factory, that is, the preloaded stack always exists; the terminal can also be successfully opened when the terminal is powered on or the terminal is unlocked successfully Create later; it can also be created after the application preload event is triggered (before the target application is determined) and so on. In an embodiment, pre-loading an application interface corresponding to the target application based on a pre-created pre-loaded active window stack includes: determining whether a pre-created pre-loaded active window stack exists; if not, according to the pre-built Set a rule to create a preloaded active window stack; preload an application interface corresponding to the target application based on the created preloaded active window stack. The advantage of this setting is that after determining the target application to be preloaded, it is determined whether the preloaded stack exists. If it exists, it does not need to be created. If it does not exist, it is created to save system resources. It can be understood that when the target application contains multiple, that is, when multiple applications need to be preloaded continuously in a short period of time, before the first target application starts to load, the preloading Stack has been created, then the first Before the two target applications start to load, the pre-loaded stack exists, so you don't need to make the above judgment.

In the embodiment of the present application, the specific process of the application interface corresponding to the preload target application based on the preloaded stack is not limited. For example, the application interface may be drawn and displayed based on the size of the preloaded stack.

In some embodiments, pre-loading an application interface corresponding to the target application based on a pre-created pre-loaded active window stack includes: creating a target process corresponding to the target application; in a pre-created pre-loaded active window A task stack corresponding to the target application is created in the stack; an active window corresponding to the target application is started in the task stack based on the target process; and a target application corresponding to the target application is drawn and displayed based on the started active window. Application interface. The advantage of this setting is that it can draw and display the application interface of the target application based on the preloaded active window stack outside the screen coordinate range, without interfering with the operation and display of the foreground application, ensuring system stability, and effectively improving The speed at which the target application starts. When the target process is created, the initialization process of the target process can also be included. During the execution of the above steps, pre-loading of other resources may also be involved, such as application service startup, memory allocation, file content reading, and network data acquisition. The embodiments of this application do not limit the pre-loading process of other resources.

In some embodiments, the method further includes: sending a fake focus notification to the target application, and maintaining a continuous drawing and display of an application interface corresponding to the target application within a third preset time period based on the fake focus notification. Update. The advantage of this setting is that it can complete the drawing and display of the application interface when the target application obtains focus and is visible to the system, improving the completion of preloading without affecting the focus of the foreground application. The focus in the embodiments of the present application is also referred to as input focus, and the fake focus and the focus corresponding to the foreground application are independent of each other. Generally, for the current Android system, the focus is unique. For example, input operations such as touch only take effect on the focus. For the input of focus information, the system and the application are consistent. Once the system changes the input focus information, it will Send information to the application that the input focus information has changed, this way to ensure that the system and application side input focus information is consistent. In the embodiment of the present application, the purpose of falsifying the focus on the application side is achieved by separating the system side and the application side from inputting focus information. In the embodiment of the present application, the focus notification for the preloaded application is forged, so that the preloaded application has focus information, but the focus information on the system side is still correct. Such processing can enable the preloaded application to draw normally and achieve the purpose of all loading. The focus exists on the system side and the application side. It can be considered as the server and the client. The system side records the application that has the focus, and the application side saves a flag to indicate whether it has focus. The timing of falsifying the input focus may be that when the Android window system adds a new window and the focus needs to be updated, a fake focus notification is generated and sent. The method of falsifying the focus may be a method of calling the client side of the window to change the focus of the window, so that the window acquires focus. In one embodiment, a fake focus notification can be sent based on the Binder mechanism. The Binder mechanism is the most commonly used method for inter-process communication of the Android system, and adopts a c / s architecture, that is, a client / service architecture.

In the embodiment of the present application, the first preset time period, the second preset time period, and the third preset time period may be designed according to actual conditions. For example, the third preset time period may be within a fixed time range after starting preloading. , It can also be the period from the start of preloading to the completion of preloading. In some embodiments, the length of the third preset time period includes a playing time of starting an advertisement or starting an animation in the target application. During the startup of some applications, some advertisements or animations are usually played. The duration of advertisements or animations usually ranges from 3 seconds to more than ten seconds. During the playback of advertisements or animations, the user may not be able to perform any operation, only Can wait for the playback to finish, wasting valuable time of the user. The advantage of this embodiment of the present application is that it can play the startup advertisement or startup animation off the screen before the target application is launched. When the target application is launched, it can directly enter the main page of the application or other user-operable Interface to further advance the operable time point of the target application and reduce waiting time.

In some embodiments, after pre-loading an application interface corresponding to the target application based on a pre-created pre-loaded active window stack, the method further includes: when receiving a running instruction of the target application, loading the pre-load An application interface corresponding to a target application corresponding to the running instruction included in the active window stack is migrated to the display screen for display. The advantage of this setting is that when the target application really needs to be started, the already drawn application interface is directly migrated to the display screen for display, which can achieve rapid switching of the application interface and improve startup speed.

In some embodiments, the migrating an application interface corresponding to a target application corresponding to the running instruction included in the preloading activity window stack to the display screen for display includes: moving the preloading activity The task stack corresponding to the target application corresponding to the run instruction included in the window stack is migrated to the top of the application active window stack; the size information, configuration information, and visibility of the task stack are updated to achieve the target application The corresponding application interface is displayed on the display screen. The advantage of this setting is that it can ensure the stability of the interface migration process and ensure that no issues such as screen freezes, black screens, or slow migration speeds occur during the recovery process.

For some terminals, such as mobile phones and tablet computers, in order to facilitate the use of the user, the display mode of the display screen usually includes vertical screen display and horizontal screen display. Many applications use vertical screen display by default, and some applications By default, the program is displayed in landscape mode (such as some online games). During the use of the terminal, some applications will switch between horizontal and vertical display as the user holds the terminal. In some embodiments of the present application, a boundary coordinate corresponding to the preloaded active window stack is (H, 0, 2H, H), a coordinate system corresponding to the boundary coordinate is a system coordinate, and an origin of the system coordinate is In the upper left corner of the display screen, H is the length of the long side of the display area of the display screen. That is, the side corresponding to H is the largest side of the display area of the display screen, which is the height of the display screen when displayed in the vertical screen and the width of the display screen when displayed in the horizontal screen. The purpose of this setting is to consider the horizontal screen of the display screen, the horizontal screen display of pre-loaded applications, and the normal display of some applications. FIG. 2 is a schematic diagram of a relative position relationship between a preloaded active window stack and a display screen display area according to an embodiment of the present application. As shown in FIG. 2, the display screen is in the vertical screen mode, the origin of the terminal system coordinates is the left vertex (0,0) of the display screen 201, the width direction of the display screen 201 is the X axis, and the height direction is the Y axis. The boundary coordinates corresponding to loading Stack202 are (H, 0, 2H, H), where H is the screen height, that is, the area within the solid rectangle on the left is the main screen display area, and the area within the rectangle on the right dashed line is the preload display area. FIG. 3 is a schematic diagram of a relative position relationship between another preloaded active window stack and a display screen display area according to an embodiment of the present application. As shown in Figure 3, the display screen is in landscape mode, the origin of the terminal system coordinates is the left vertex (0,0) of the display screen 301, the height direction of the display screen 301 is the X axis, and the width direction is the Y axis. The boundary coordinates corresponding to loading Stack202 are (H, 0, 2H, H), where H is the screen height, that is, the area within the solid rectangle on the left is the main screen display area, and the area within the rectangle on the right dashed line is the preload display area.

The reason for the pre-loaded Stack boundary is so:

The horizontal axis of the upper left corner is H, which is to prevent the display screen (also called the home screen) from being displayed on the pre-loaded application interface when the screen is horizontal, because the home screen has a horizontal screen mode in addition to the vertical screen mode. In order to prevent part of the preloaded application from being displayed in the main screen display area, the horizontal coordinate of the upper left corner of the rectangular area corresponding to the preloaded stack is set to the screen height.

The vertical coordinate of the upper left corner is 0, so that the preloaded application can correctly calculate the height of the status bar. In order to better design the user interface (UI), the Android application will customize the top status bar. If the corresponding vertical coordinate is not equal to 0, the height of the status bar may be wrong.

The horizontal axis of the lower right corner is 2H (2 times the screen height), that is, the width of the rectangle corresponding to the preloaded stack = screen height, so that the size of the preloaded stack can include the horizontal screen application when preloaded (that is, the application interface is horizontal screen) Display mode application).

The vertical coordinate of the lower right corner is H, that is, the height of the rectangle corresponding to the pre-loaded Stack = screen height, so that the size of the pre-loaded Stack can include the vertical screen application when pre-loaded.

Based on the above reasons, the size of the preloaded stack is set to (H, 0, 2H, H).

In addition, FIG. 4 is a schematic diagram of an application interface migration provided by an embodiment of the present application. As shown in FIG. 4, upon receiving a running instruction of a target application, an application corresponding to the target application included in the active window stack is preloaded. The interface 401 is migrated to the display screen 201 for display. In one embodiment, the task to which the preloaded application interface belongs is migrated to the top of the application activity window Stack, and the size information, configuration information, and visibility of the task are updated to thereby apply the interface. Able to display normally on the display screen.

FIG. 5 is a schematic flowchart of another application preloading method according to an embodiment of the present application. The method includes the following steps:

Step 501: Collect a current sample when it is detected that an application preload event is triggered.

Step 502: Input the current sample into a first-type prediction model based on an RNN network, and determine a target candidate application set according to an output result of the first-type prediction model.

Step 503: Input the current sample into a second type prediction model based on the DNN network corresponding to the target candidate application set, and determine a target application to be preloaded in the target candidate application set according to an output result of the second type prediction model program.

Step 504: Pre-load an application interface corresponding to the target application based on a pre-created pre-loaded active window stack, wherein a boundary coordinate corresponding to the pre-loaded active window stack is outside a coordinate range of the display screen.

Step 505: When the running instruction of the target application is received, the application interface of the target application corresponding to the running instruction contained in the preloaded active window stack is migrated to a display screen for display.

The application preloading method provided in the embodiment of the present application can use the same training samples to separately train the RNN network and the DNN network. When training the RNN network, the samples corresponding to the training samples are labeled as candidate application sets, and in When training a DNN network, the samples corresponding to the training samples are labeled as specific applications. This saves time during the training and sample collection stages. When the application preload event is triggered, the current sample is first input into the first type of prediction model based on the RNN network, the target candidate application set is predicted, and then the current sample is input into the second type of prediction model based on the DNN network. The application to be preloaded is predicted, and the off-screen preloading method is used to avoid the interference of the preloading process on the foreground application. When the user actually starts the target application, the application interface that is preloaded offscreen can be directly migrated. Go to the display screen for display, effectively improving the startup speed of the target application.

FIG. 6 is a schematic flowchart of another application preloading method according to an embodiment of the present application. The method includes:

Step 601: Collect the historical use time-series correlation sequence of the sample application in the first preset time period as the first training sample, and collect historical state characteristic information of the terminal when the sample application is used as the second training sample.

Wherein, the sample mark of the first training sample includes a candidate application set to which an application used after the sample application belongs; the sample mark of the second training sample includes a candidate used after the sample application application.

For example, when collecting the first training sample and the second training sample, the candidate application set to which they belong may not be distinguished temporarily, and only the sample mark of the second training sample is recorded, that is, the application used after the sample application is recorded. program.

Exemplarily, the length of the first preset time period may be, for example, 2 months. After the training sample is collected, the numbers are sorted according to the frequency of each application used within the 2 months. Assign the highest ID number to the most frequently used application, and the lowest ID number to the least frequently used application. Assuming a total of 100 candidate applications, the final number is 1 to 100, and then the 100 IDs are divided into n types of applications, such as 4 categories: most commonly used, commonly used, infrequently used, rarely used, 25 in each category . That is, it is divided into 4 candidate application sets according to frequency, and each candidate application set contains 25 candidate application programs.

Step 602: Input a first training sample and a corresponding sample mark into a first preset model to train the first preset model, and use the trained model as a first-type prediction model.

Exemplarily, the first preset model may be a machine learning model based on an RNN network.

Step 603: For each candidate application set, input a second training sample corresponding to the current candidate application set and a corresponding sample tag into a second preset model to train the second preset model, and train the The second preset model is used as a second type prediction model corresponding to the current candidate application set.

Exemplarily, the second preset model may be a machine learning model based on a DNN network. In this step, the above example is used as an example for illustration. Four second-type prediction models are obtained, that is, each of the four candidate application sets has its own corresponding second-type prediction model.

Step 604: Collect the first current sample and the second current sample when it is detected that the application preload event is triggered.

Step 605: The first current sample is input to a first-type prediction model, and a target candidate application set is determined according to an output result of the first-type prediction model.

Exemplarily, the determined target candidate application set is the most commonly used candidate application set.

Step 606: Determine the corresponding second-type prediction model according to the target candidate application set, input the second current sample to the second-type prediction model, and determine the target application to be preloaded according to the output result of the second-type prediction model.

Exemplarily, the second current sample is input into a second-type prediction model corresponding to the most commonly used candidate application set, and one or more target applications are selected from 25 applications according to the model output result.

Step 607: Pre-load an application interface corresponding to the target application based on a pre-created pre-loaded active window stack.

The boundary coordinates corresponding to the preloaded active window stack are located outside the coordinate range of the display screen.

Step 608: When the running instruction of the target application is received, the application interface of the target application corresponding to the running instruction contained in the preloaded active window stack is migrated to a display screen for display.

In the application preloading method provided in the embodiment of the present application, a terminal collects a first training sample and a second training sample in a historical period, and trains a first type prediction model for predicting a target candidate application set based on the first training sample. The second training sample is trained to obtain the second type of prediction model used to predict the application to be preloaded, so as to achieve the purpose of sample collection and model training locally at the terminal, and reduce data transmission, making the sample collection and prediction model update more timely. At the same time, it can save traffic and can be applied to mobile terminals such as mobile phones. When the application preload event is triggered, two types of prediction models are used to perform segmented prediction to improve the prediction accuracy and accuracy of the application. After predicting the application to be preloaded, the off-screen preloading method is used to avoid the interference of the preloading process to the foreground application. When the user actually launches the target application, the preloaded application can be directly preloaded offscreen. The interface is migrated to the display screen for display, which effectively improves the startup speed of the target application.

FIG. 7 is a structural block diagram of an application preloading device according to an embodiment of the present application. The device may be implemented by software and / or hardware, and is generally integrated in a terminal. The application preloading method may be performed to perform application preloading. load. As shown in Figure 7, the device includes:

An application set determining module 701 configured to determine a target candidate application set by using a first-type prediction model when an application preload event is detected to be triggered;

A target application determining module 702 configured to determine a target application to be preloaded in the target candidate application set by using a second type prediction model corresponding to the target candidate application set;

The preloading module 703 is configured to preload the target application program.

The application preloading device provided in the embodiment of the present application, when detecting that an application preloading event is triggered, uses a first type prediction model to determine a target candidate application set, and then uses a second type prediction model corresponding to the target candidate application set to determine a target. The target application to be preloaded included in the candidate application set is used to preload the target application. By adopting the above technical solution, two different types of prediction models can be used for staged prediction, which improves the accuracy of application prediction.

In an embodiment, the application set determination module 701 is configured to:

Obtaining a current use time-series association sequence of a foreground application, where the current use time-series association sequence includes a sequence formed by the foreground application and at least one application used before the foreground application in chronological order;

The current used time-series correlation sequence is input into a first-type prediction model, and a target candidate application set is determined according to an output result of the first-type prediction model.

In an embodiment, the device further includes:

A first training sample acquisition module is configured to collect a history of a sample application in a first preset time period using a time-series association sequence as a first training sample before the detection that an application preload event is triggered. A sample tag of a training sample includes a candidate application set to which an application program used after the sample application program belongs;

A first model training module configured to input the first training sample and a corresponding sample marker into a first preset model to train the first preset model, and use the trained model as a first class Forecasting model.

In one embodiment, the target application determination module 702 is configured to:

Obtain the current state characteristic information of the terminal;

Inputting the current state characteristic information into a second type prediction model corresponding to the target candidate application set, and determining a target to be preloaded in the target candidate application set according to an output result of the second type prediction model application.

In an embodiment, the device further includes:

A second training sample acquisition module is configured to collect historical state characteristics of the terminal when a sample application in a current candidate application set is used within a second preset time period before the detection that an application preload event is triggered Information as the second training sample, and the sample tag of the second training sample includes an application program used after the sample application program;

A second model training module configured to input the second training sample and a corresponding sample marker into a second preset model to train the second preset model and to train the second preset model after training As a second type prediction model corresponding to the current candidate application set.

In an embodiment, the division rule of the candidate application set includes division according to an application type or division according to a usage frequency of an application during a historical usage period.

In one embodiment, the preloading module 703 is configured to:

An application interface corresponding to the target application is preloaded based on a pre-created preloaded active window stack, wherein a boundary coordinate corresponding to the preloaded active window stack is outside a coordinate range of a display screen.

In one embodiment, the preloading module 703 is configured to:

Creating a target process corresponding to the target application;

Creating a task stack corresponding to the target application in a pre-created preloading active window stack;

Starting an active window corresponding to the target application in the task stack based on the target process;

An application interface corresponding to the target application is drawn and displayed based on the activated active window.

In an embodiment, the device is further configured to:

When a running instruction of the target application is received, an application interface corresponding to the target application corresponding to the running instruction included in the preloaded active window stack is migrated to the display screen for display.

In an embodiment, the device is further configured to:

Migrating the task stack corresponding to the target application corresponding to the running instruction contained in the preloaded active window stack to the top of the application active window stack;

Update the size information, configuration information, and visibility of the task stack to enable the application interface corresponding to the target application to be displayed on the display screen.

In an embodiment, the device is further configured to:

Sending a fake focus notification to the target application, and maintaining a continuous drawing and display update of an application interface corresponding to the target application for a third preset time period based on the fake focus notification.

In one embodiment, the preloading module 703 is configured to:

When there are multiple target applications, pre-load the multiple target applications in sequence; or

When there are multiple target applications, at least two target applications among the multiple target applications are preloaded simultaneously.

An embodiment of the present application further provides a storage medium including computer-executable instructions, where the computer-executable instructions are used to execute an application program preloading method when executed by a computer processor, the method includes:

When it is detected that the application preload event is triggered, the first type of prediction model is used to determine the target candidate application set;

Determining a target application to be preloaded included in the target candidate application set by using a second type prediction model corresponding to the target candidate application set;

Pre-load the target application.

Storage medium-any one or more types of memory devices or storage devices. The term "storage medium" is intended to include: installation media, such as Compact Disc Read-Only Memory (CD-ROM), floppy disks, or magnetic tape devices; computer system memory or random access memory, such as dynamic random access memory Access Memory (Dynamic Random Access Memory, DRAM), Double Rate Random Access Memory (DDR Random Access Memory, DDR RAM), Static Random Access Memory (Static Random Access Memory, SRAM), and extended data output random storage Access Memory (Extended Data, Random-Access Memory, EDO, RAM), Rambus Random-Access Memory (Random-Access Memory, RAM), etc .; Non-volatile memory, such as flash memory, magnetic media (such as hard disk or Optical storage); registers or other similar types of memory elements, etc. The storage medium may further include other types of memory or a combination thereof. In addition, the storage medium may be located in a first computer system in which the program is executed, or may be located in a different second computer system connected to the first computer system through a network such as the Internet. The second computer system may provide program instructions to the first computer system, and the first computer system is configured to execute the program instructions. The term "storage medium" may include two or more storage media that may reside in different locations, such as in different computer systems connected through a network. The storage medium may store program instructions (for example, embodied as a computer program) executable by one or more processors.

Certainly, a storage medium including computer-executable instructions provided in the embodiments of the present application is not limited to the application preloading operation described above, and may also execute the application provided by any embodiment of the present application Related operations in the preload method.

The embodiment of the present application provides a terminal, and the terminal may integrate the application preloading device provided by the embodiment of the present application. FIG. 8 is a schematic structural diagram of a terminal according to an embodiment of the present application. The terminal 800 may include: a memory 801, a processor 802, and a computer program stored on the memory 801 and executable by the processor 802. When the processor 802 executes the computer program, the application program according to the embodiment of the present application is implemented. Preload method.

The terminal provided in the embodiment of the present application can use two different types of prediction models to perform staged prediction, thereby improving the accuracy of application prediction.

FIG. 9 is a schematic structural diagram of another terminal according to an embodiment of the present application. The terminal may include: a housing (not shown in the figure), a memory 901, a central processing unit (CPU) 902 (also referred to as processing). Device), a circuit board (not shown in the figure), and a power supply circuit (not shown in the figure). The circuit board is disposed in a space surrounded by the housing; the CPU 902 and the memory 901 are provided on the circuit board; and the power supply circuit is provided as a plurality of circuits or devices of the terminal The memory 901 is configured to store executable program code; the CPU 902 reads the executable program code stored in the memory 901 to run a computer program corresponding to the executable program code to implement the following step:

When it is detected that the application preload event is triggered, the first type of prediction model is used to determine the target candidate application set;

Determining a target application to be preloaded included in the target candidate application set by using a second type prediction model corresponding to the target candidate application set;

Pre-load the target application.

The terminal further includes: a peripheral interface 903, a radio frequency (RF) circuit 905, an audio circuit 906, a speaker 911, a power management chip 908, an input / output (I / O) subsystem 909, and other The input / control device 910, the touch screen 912, other input / control devices 910, and the external port 904, these components communicate through one or more communication buses or signal lines 907.

It should be understood that the illustrated terminal 900 is only an example of the terminal, and the terminal 900 may have more or fewer components than those shown in the figure, may combine two or more components, or may have Different component configurations. The one or more components shown in the figures may be implemented in hardware, software, or a combination of hardware and software including one or more signal processing and / or application specific integrated circuits.

The following describes the terminal used for application preloading provided in this embodiment in detail. The terminal uses a mobile phone as an example.

Memory 901, which can be accessed by CPU 902, peripheral interface 903, etc. The memory 901 can include high-speed random access memory, and can also include non-volatile memory, such as one or more disk storage devices, flash memory Devices, or other volatile solid-state storage devices.

A peripheral interface 903, which can connect input and output peripherals of the device to the CPU 902 and the memory 901.

I / O subsystem 909, which can connect input / output peripherals on the device, such as touch screen 912 and other input / control devices 910, to peripheral interface 903. The I / O subsystem 909 may include a display controller 9091 and one or more input controllers 9092 for controlling other input / control devices 910. Among them, one or more input controllers 9092 receive electrical signals from or send electrical signals to other input / control devices 910. Other input / control devices 910 may include physical buttons (press buttons, rocker buttons, etc.) ), Dial, slide switch, joystick, click wheel. It is worth noting that the input controller 9092 may be connected to any of the following: a keyboard, an infrared port, a universal serial bus (Universal Serial Bus, USB) interface, and a pointing device such as a mouse.

A touch screen 912, which is an input interface and an output interface between a user terminal and a user, and displays a visual output to the user. The visual output may include graphics, text, icons, videos, and the like.

The display controller 9091 in the I / O subsystem 909 receives an electric signal from the touch screen 912 or sends an electric signal to the touch screen 912. The touch screen 912 detects a contact on the touch screen, and the display controller 9091 converts the detected contact into interaction with a user interface object displayed on the touch screen 912, that is, realizes human-computer interaction. The user interface object displayed on the touch screen 912 may be running Icons for games, icons connected to the appropriate network, etc. It is worth noting that the device may also include a light mouse, which is a touch-sensitive surface that does not display visual output, or an extension of the touch-sensitive surface formed by a touch screen.

The RF circuit 905 is configured to establish communication between a mobile phone and a wireless network (that is, a network side), and realize data reception and transmission of the mobile phone and the wireless network. For example, send and receive text messages, e-mail, and so on. In an embodiment, the RF circuit 905 receives and sends RF signals. The RF signals are also referred to as electromagnetic signals. The RF circuit 905 converts electrical signals into electromagnetic signals or converts electromagnetic signals into electrical signals, and communicates with the communication network through the electromagnetic signals. As well as other devices. RF circuit 905 may include known circuits for performing these functions, including, but not limited to, antenna systems, RF transceivers, one or more amplifiers, tuners, one or more oscillators, digital signal processors, codecs (COder-DECoder, CODEC) chipset, Subscriber Identity Module (SIM), and so on.

The audio circuit 906 is configured to receive audio data from the peripheral interface 903, convert the audio data into an electrical signal, and send the electrical signal to the speaker 911.

The speaker 911 is configured to restore a voice signal received by the mobile phone from the wireless network through the RF circuit 905 to a sound and play the sound to a user.

The power management chip 908 is configured to provide power and power management for the hardware connected to the CPU 902, the I / O subsystem, and peripheral interfaces.

The application preloading device, storage medium, and terminal provided in the foregoing embodiments can execute the application preloading method provided in any embodiment of the present application, and have corresponding function modules and effects for executing the method. For technical details not described in detail in the foregoing embodiments, reference may be made to an application preloading method provided in any embodiment of the present application.

Claims (20)

1. An application preloading method includes:

   When it is detected that the application preload event is triggered, the first type of prediction model is used to determine the target candidate application set;

   Determining a target application to be preloaded included in the target candidate application set by using a second type prediction model corresponding to the target candidate application set;

   Pre-load the target application.
2. The method according to claim 1, wherein the determining a target candidate application set by using the first-type prediction model comprises:

   Obtaining a current use time-series association sequence of a foreground application, where the current use time-series association sequence includes a sequence formed by the foreground application and at least one application used before the foreground application in chronological order;

   Inputting the currently used time-series correlation sequence into the first-type prediction model, and determining the target candidate application set according to an output result of the first-type prediction model.
3. The method according to claim 2, before the detecting that an application preload event is triggered, further comprising:

   Collecting a history of a sample application in a first preset time period using a time-series association sequence as a first training sample, and the sample tag of the first training sample includes a candidate to which an application used after the sample application belongs Application collection

   Inputting the first training sample and the corresponding sample mark into a first preset model to train the first preset model, and using the trained model as the first type prediction model.
4. The method according to claim 1 or 2, wherein the determining a target application to be preloaded included in the target candidate application set by using a second type prediction model corresponding to the target candidate application set comprises:

   Obtain the current state characteristic information of the terminal;

   Inputting the current state characteristic information into a second type prediction model corresponding to the target candidate application set, and determining a target to be preloaded in the target candidate application set according to an output result of the second type prediction model application.
5. The method according to claim 4, before the detecting that an application preload event is triggered, further comprising:

   Collect historical state characteristic information of the terminal when a sample application in the current candidate application set is used within a second preset time period, as a second training sample, and the sample mark of the second training sample includes the sample application A program or an application that is used after the sample application;

   Inputting the second training sample and a corresponding sample tag into a second preset model to train the second preset model, and using the trained second preset model as the current candidate application set corresponding The second type of prediction model.
6. The method according to claim 1, wherein the division rule of the candidate application set comprises division according to an application type or division according to a usage frequency of an application during a historical usage period.
7. The method according to claim 1, wherein the pre-loading the target application comprises:

   An application interface corresponding to the target application is preloaded based on a pre-created preloaded active window stack, wherein a boundary coordinate corresponding to the preloaded active window stack is outside a coordinate range of a display screen.
8. The method according to claim 7, wherein the pre-loading an application interface corresponding to the target application based on a pre-created pre-loaded active window stack comprises:

   Creating a target process corresponding to the target application;

   Creating a task stack corresponding to the target application in a pre-created preloading active window stack;

   Starting an active window corresponding to the target application in the task stack based on the target process;

   An application interface corresponding to the target application is drawn and displayed based on the activated active window.
9. The method according to claim 7, after the pre-loading an application interface corresponding to the target application based on a pre-created pre-loaded active window stack, further comprising:

   When a running instruction of the target application is received, an application interface corresponding to the target application corresponding to the running instruction included in the preloaded active window stack is migrated to the display screen for display.
10. The method according to claim 9, wherein migrating an application interface corresponding to a target application corresponding to the running instruction contained in the preloaded active window stack to the display screen for display comprises:

    Migrating the task stack corresponding to the target application corresponding to the running instruction contained in the preloaded active window stack to the top of the application active window stack;

    Update the size information, configuration information, and visibility of the task stack to enable the application interface corresponding to the target application to be displayed on the display screen.
11. The method according to claim 7 or 8, further comprising:

    Sending a fake focus notification to the target application, and maintaining a continuous drawing and display update of an application interface corresponding to the target application for a third preset time period based on the fake focus notification.
12. The method according to claim 1, wherein the pre-loading the target application comprises:

    When there are multiple target applications, pre-load the multiple target applications in sequence; or

    When there are multiple target applications, at least two target applications among the multiple target applications are preloaded simultaneously.
13. An application preloading device includes:

    The application set determination module is configured to determine a target candidate application set by using a first-type prediction model when an application preload event is detected to be triggered;

    A target application determination module configured to determine a target application to be preloaded in the target candidate application set by using a second type prediction model corresponding to the target candidate application set;

    The preloading module is configured to preload the target application.
14. The apparatus according to claim 13, wherein the application set determination module is configured to:

    Obtaining a current use time-series association sequence of a foreground application, where the current use time-series association sequence includes a sequence formed by the foreground application and at least one application used before the foreground application in chronological order;

    Inputting the currently used time-series correlation sequence into the first-type prediction model, and determining the target candidate application set according to an output result of the first-type prediction model.
15. The apparatus according to claim 14, further comprising:

    A first training sample collection module is configured to collect a historical use time-series correlation sequence of a sample application within a first preset period of time as a first training sample, and a sample tag of the first training sample is included in the sample application The set of candidate applications to which the applications to be used later belong;

    The first model training module is configured to input the first training sample and a corresponding sample mark into a first preset model to train the first preset model, and use the trained model as the first A class of predictive models.
16. The apparatus according to claim 13 or 14, wherein the target application determination module is configured to:

    Obtain the current state characteristic information of the terminal;

    Inputting the current state characteristic information into a second type prediction model corresponding to the target candidate application set, and determining a target to be preloaded in the target candidate application set according to an output result of the second type prediction model application.
17. The apparatus according to claim 16, further comprising:

    A second training sample collection module is configured to collect historical state feature information of the terminal when a sample application in a current candidate application set is used within a second preset time period, as a second training sample, the second training The sample tag of the sample includes the sample application or an application used after the sample application;

    A second model training module configured to input the second training sample and a corresponding sample marker into a second preset model to train the second preset model and to train the second preset model after training As a second type prediction model corresponding to the current candidate application set.
18. The apparatus according to claim 13, the preloading module is configured to:

    An application interface corresponding to the target application is preloaded based on a pre-created preloaded active window stack, wherein a boundary coordinate corresponding to the preloaded active window stack is outside a coordinate range of a display screen.
19. A computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the method according to any one of claims 1-12 is implemented.
20. A terminal includes a memory, a processor, and a computer program stored on the memory and executable by the processor. When the processor executes the computer program, the processor implements the device according to any one of claims 1-12. The method described.

Images