WO2019062462A1 - Application control method and apparatus, storage medium and electronic device - Google Patents

Abstract

Disclosed are an application control method and apparatus, a storage medium and an electronic device. The application control method is applied to the electronic device, and comprises: acquiring operating parameters of the electronic device at each sampling time point in a historical time period; according to the sampling time point and the operating parameters, generating a training sample; using the training sample to train a pre-set Bayesian model; and based on the trained Bayesian model, controlling a background application in the electronic device.

Description

Application control method, device, storage medium, and electronic device

This application claims the priority of the Chinese Patent Application, filed on Sep. 30, 2017, filed Jan. In this application.

Technical field

The present application relates to the field of computer technologies, and in particular, to an application control method, apparatus, storage medium, and electronic device.

Background technique

With the development of technology, mobile terminals such as smartphones and tablet PCs (PADs) have become indispensable devices in user life.

At present, there are more and more applications installed on the terminal. When the user finishes using the application in the terminal, the user usually performs operations such as switching to a new application, returning to the main interface, or locking the screen. The program is switched to the background, these applications in the background will continue to run, for example, exchange data with the server, listen to user actions, etc.; during the running process, it will continue to occupy system resources, for example, occupy system memory, consume data traffic, consume Terminal power, etc. In order to prevent the used application from continuing to occupy system resources, it is usually necessary to clean up the background application, such as selecting the application that consumes more memory according to the amount of memory consumed by the application, or according to the application set at the factory. Program priority, clean up low-priority applications, and so on, but these cleanup methods are relatively rigid, unable to flexibly determine which applications can be cleaned, and it is difficult to effectively improve system resources.

Summary of the invention

The embodiments of the present application provide an application control method, device, storage medium, and electronic device, which can flexibly clean background applications and effectively improve system resources.

The embodiment of the present application provides an application control method, which is applied to an electronic device, and includes:

Obtaining an operating parameter of the electronic device at each sampling time point in the historical period, where the operating parameter includes a foreground application, a remaining power, a screened duration, a charging connection state, and a network connection state;

Generating a training sample according to the sampling time point and the running parameter;

Training the preset Bayesian model with the training sample;

The background application in the electronic device is controlled based on the trained Bayesian model.

The embodiment of the present application further provides an application control device, which is applied to an electronic device, and includes:

An acquiring module, configured to acquire an operating parameter of the electronic device at each sampling time point in a historical period, where the operating parameter includes a foreground application, a remaining power, a screened duration, a charging connection state, and a network connection state;

a generating module, configured to generate a training sample according to the sampling time point and the running parameter;

a training module, configured to train the preset Bayesian model by using the training sample;

And a control module, configured to control a background application in the electronic device based on the trained Bayesian model.

The embodiment of the present application further provides a storage medium, where the storage medium stores a plurality of instructions, and the instructions are adapted to be loaded by a processor to execute any one of the application control methods described above.

The embodiment of the present application further provides an electronic device, including a processor and a memory, the processor is electrically connected to the memory, the memory is used to store instructions and data, and the processor is used in any one of the foregoing The steps in the application control method described.

DRAWINGS

The technical solutions and other advantageous effects of the present application will be apparent from the detailed description of the embodiments of the present application.

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

FIG. 2 is another schematic flowchart of an application control method according to an embodiment of the present application.

FIG. 3 is a flowchart of training a Bayesian model provided by an embodiment of the present application.

FIG. 4 is a schematic diagram of a frame of a Bayesian model provided by an embodiment of the present application.

FIG. 5 is a schematic structural diagram of an application control apparatus according to an embodiment of the present application.

FIG. 6 is a schematic structural diagram of a generation module according to an embodiment of the present application.

FIG. 7 is a schematic structural diagram of a generating submodule according to an embodiment of the present disclosure.

FIG. 8 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed ways

The technical solutions in the embodiments of the present application are clearly and completely described in the following with reference to the drawings in the embodiments of the present application. It is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments obtained by a person skilled in the art based on the embodiments of the present application without creative efforts are within the scope of the present application.

Embodiments of the present application provide an application control method, apparatus, storage medium, and electronic device.

An application control method is applied to an electronic device, comprising: acquiring an operating parameter of the electronic device at each sampling time point in a historical time period, where the operating parameter includes a foreground application, a remaining power, a screened duration, a charging connection state, and a network connection. a training sample is generated according to the sampling time point and the running parameter; the preset Bayesian model is trained by using the training sample; and the background application in the electronic device is controlled based on the trained Bayesian model.

In some embodiments, the generating the training sample according to the sampling time point and the operating parameter comprises:

Determining a sampling date type and a sampling period according to the sampling time point;

Determining a preset power range to which the remaining power belongs, and determining a preset duration range to which the bright screen duration belongs;

The training samples are generated according to the sampling date type, the sampling period, the preset power range, the preset duration range, the foreground application, the charging connection state, and the network connection state.

In some embodiments, the generating the training sample according to the sampling date type, the sampling period, the preset power range, the preset duration range, the foreground application, the charging connection state, and the network connection state, includes:

Obtain a target prediction application;

Determining, according to the sampling time point, a previous switching application and a next switching application of the foreground application from the operating parameters;

Determining a predicted value of the target prediction application according to the sampling time point, a next switching application, and a foreground application;

A training sample is generated according to the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, the target prediction application, and the predicted value.

In some embodiments, the determining, according to the sampling time point, the next switching application, and the foreground application, the predicted value of the target prediction application, including:

Calculating a difference between a sampling time point of the next switching application and a sampling time point of the foreground application;

Determining whether the target prediction application is the next handover application, and whether the difference does not exceed a preset duration;

If yes, determining the predicted value of the target prediction application as a first preset value;

If not, the predicted value of the target prediction application is determined as a second preset value.

In some embodiments, the according to the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection status, the network connection status, the target prediction application, and the predicted value. Generate training samples, including:

Obtaining, respectively, the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, and the feature value corresponding to the target prediction application;

A training sample is generated based on the feature value and the predicted value.

In some embodiments, the feature value comprises (q ₁ , q ₂ ... q _m ), the predicted value comprises j1 and j2, and the training of the preset Bayesian model is performed by using the training sample, including :

And inputting the predicted value into the first preset formula to obtain a probability of the corresponding predicted value, where the first preset formula is:

Where N(j1) represents the number of occurrences of event j1, N(j2) represents the number of occurrences of event j2, and P(j2) represents the probability of occurrence of event j2;

And inputting the feature value and the predicted value into the second preset formula to obtain a probability of the corresponding feature value and the predicted value, where the second preset formula is:

Where 1 ≤ _i ≤ m, P(q _i | j2) represents the probability that the event q _i occurs on the premise that the event j2 occurs, and N(q _i , j2) represents the number of times the events q _i and j2 occur simultaneously.

In some embodiments, the training based Bayesian model controls a background application in the electronic device, including:

Get background application cleanup instructions;

Obtaining, according to the background application cleanup instruction, a background application of the electronic device, and current running parameters;

Calculate the cleanable rate of each background application by using the trained Bayesian model and the current operating parameters;

The background application is closed according to the cleanup rate.

In some embodiments, the calculating the cleanable rate of each background application using the trained Bayesian model and the current operational parameters includes:

Determining a current feature value according to current operating parameters;

The current feature value is input into the third preset formula for calculation, and the cleanable rate is obtained. The third preset formula is: P

Where 1 ≤ _k ≤ m, q _k is the current eigenvalue.

In some embodiments, the closing the background application according to the cleanup rate comprises:

Select a background application whose cleaning rate is not less than the preset threshold as the target application, or select the preset number of background applications with the highest cleanup rate as the target application;

Close the target app.

As shown in FIG. 1, the application control method is applied to an electronic device, and the specific process can be as follows:

101. Obtain an operating parameter of the electronic device at each sampling time point in the historical time period, where the operating parameter includes a foreground application, a remaining power, a screened duration, a charging connection state, and a network connection state.

In this embodiment, the historical period may be manually set, such as the previous month or the first two months. The sampling time point is mainly determined according to the sampling frequency, for example, it can be sampled every minute or every two minutes, and can be expressed in the form of x minutes of x years x months x days x. Both the charging connection state and the network connection state may include both connected and unconnected situations.

In the actual application process, the running parameter may be obtained in real time, for example, the corresponding data collection operation is performed at the sampling time point, or may be obtained once, for example, the electronic device may record the historical time period in the local database in advance. The screen change data, the charge state change data, the network state change data, and the application open data are displayed, and then the operation parameters of each sampling time point can be extracted at one time according to the sampling frequency.

102. Generate a training sample according to the sampling time point and the running parameter.

For example, the foregoing step 102 may specifically include:

1-1. Determine a sampling date type and a sampling period according to the sampling time point.

In this embodiment, the sampling date type is divided into weekly, which may include weekdays and weekends. The sampling period is divided into daily, which can divide the day into 48 time periods.

1-2. Determine a preset power range to which the remaining power belongs, and determine a preset duration range to which the bright time duration belongs.

In this embodiment, the preset power range and the preset duration range may be manually set, and the preset power range may include three ranges indicating high power, medium power, and low power, for example, the high power may be 70%. -100%, the medium power can be 40%-70%, the low battery can be 0-40%, etc., the preset duration range can include three intervals indicating short, medium and long, for example, the length can be more than 10min, It can be 5-10min in length and 0-5min in short.

1-3. Generate training samples according to the sampling date type, sampling period, preset power range, preset duration range, foreground application, charging connection status, and network connection status.

For example, the foregoing steps 1-3 may specifically include:

1-3-1. Obtain the target prediction application.

In this embodiment, the target prediction application may be all applications installed in the electronic device, or may be partial applications. When it is a partial application, it may be several applications with the highest frequency of occurrence in the near future, and the specific quantity may be according to actual needs. And set.

1-3-2. Determine, according to the sampling time point, the last switching application and the next switching application of the foreground application from the operating parameters.

In this embodiment, since the foreground application obtained by each sampling in the historical period is known, the foreground application obtained before the sampling time point can be regarded as the current foreground application for the foreground application obtained by any sampling. The previous foreground application that is obtained after the sampling time point can be regarded as the next switching application of the current foreground application. Generally, different foreground applications that are closest to the current sampling time point can be taken as the previous switching application and Next switch application. In the actual operation process, all foreground applications may be sorted according to the sampling time point. For any three adjacent different foreground applications after sorting, the front foreground application may serve as the previous switching application of the intermediate foreground application, followed by The foreground application can serve as the next switching application for the intermediate foreground application.

1-3-3. Determine a predicted value of the target prediction application according to the sampling time point, the next switching application, and the foreground application.

In this embodiment, the predicted value may be an artificially set value, such as 0 and 1, where 0 indicates that the target prediction application will not switch to the foreground use in a short time, and 1 may indicate that the target prediction application will be in the Switch to the foreground for a short time. Since all the foreground applications collected in the historical period are known, the predicted value of the target prediction application can be determined according to the known foreground application and the sampling time point. In this case, the above steps 1-3-3 can be specifically include:

Calculating a difference between a sampling time point of the next switching application and a sampling time point of the foreground application;

Determining whether the target prediction application is the next handover application, and whether the difference does not exceed a preset duration;

If yes, determining the predicted value of the target prediction application as the first preset value;

If not, the predicted value of the target prediction application is determined as a second preset value.

In this embodiment, the preset duration, the first preset value, and the second preset value may be manually set, and the preset duration is mainly used to define a length of time, which may be 10 minutes, and the first preset value may be Yes 1, the second preset value can be 0. For each sampling, when the target prediction application that needs to be predicted is the next switching application, it is necessary to further analyze the length of time taken to switch from the current application to the next switching application, only when the interval duration is within the preset duration, The predicted value of the target prediction application can be set to 1, otherwise, all are set to 0.

1-3-4. Generate training samples according to the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, the target prediction application, and the predicted value. .

In this embodiment, in order to analyze the user behavior from multiple dimensions, so that the trained machine learning model is more anthropomorphic, each training sample may be composed of a plurality of known feature items and data of the tag items, the known The feature item may include the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection status, and the network connection status, etc., and the label item is mainly.

For example, the foregoing steps 1-3-4 may specifically include:

Obtaining, respectively, the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, and the feature value corresponding to the target prediction application;

A training sample is generated based on the feature value and the predicted value.

In this embodiment, since the computer program generally runs in the form of characters, the feature value can be mainly expressed in the form of Arabic numerals or letters, such as 1-10, and each feature item can also be expressed in the form of letters, such as the foreground. The application is H, the sampling date type is B, and so on. When generating the training sample, the feature value of the feature item may be directly used as a prior condition, and the predicted value of each target prediction application is used as a posterior result to generate the training sample.

It is easy to understand that the feature values corresponding to each feature item may be preset, and the feature values of different feature items may be the same or different, for example, the feature values of the foreground application and the sampling period may include 0 to 10, However, each number refers to a different meaning in different feature items. For example, for the foreground application, 0 can refer to the US group. For the sampling period, 0 can refer to the period from 0:00 to 1:00.

103. Train the preset Bayesian model with the training sample.

For example, the eigenvalues may include (q ₁ , q ₂ ... q _m ), and the predicted value may include j1 and j2.

And inputting the predicted value into the first preset formula to obtain a probability of the corresponding predicted value, where the first preset formula is:

Where N(j1) represents the number of occurrences of event j1, N(j2) represents the number of occurrences of event j2, and P(j2) represents the probability of occurrence of event j2;

And inputting the feature value and the predicted value into the second preset formula to obtain a probability of the corresponding feature value and the predicted value, where the second preset formula is:

Where 1 ≤ _i ≤ m, P(q _i | j2) represents the probability that the event q _i occurs on the premise that the event j2 occurs, and N(q _i , j2) represents the number of times the events q _i and j2 occur simultaneously.

In this embodiment, the Bayesian model can be:

Where m is the number of feature terms, q ₁ , q ₂ ... q _m are the eigenvalues of the a priori condition, q _i is the eigenvalue corresponding to the i-th feature item, and J is the predicted value of the target prediction application. To simplify the calculation, assuming that q ₁ , q ₂ ... q _m are independent of each other, then

Thus the Naive Bayes classifier model is obtained:

J _MAX = arg max P(J|q ₁ , q ₂ ... q _m )=arg maxP(q ₁ |J)P(q ₂ |J)...P(q _m |J), where J can represent j1 or j2 The probability value of each feature item is the statistical probability of the number of occurrences, that is, the above formula:

Where j1 is the first preset value and j2 is the second preset value. It is easy to know that the process of training the Bayesian model is the process of probability and statistics, that is, after training the Bayesian model, the probability values of different eigenvalues in each feature item can be obtained, such as P(q ₁ ), P( q ₁ |j2).

104. Control the background application in the electronic device based on the trained Bayesian model.

For example, the foregoing step 104 may specifically include:

2-1. Obtain the background application cleanup command.

In this embodiment, the background application cleanup instruction may be automatically generated by the electronic device, for example, when the memory occupancy reaches a certain limit, or the power is insufficient, or the running speed is too slow, the background application cleanup instruction is generated, and of course, the background application is cleaned up. The instructions may also be generated manually by the user. For example, the user may generate the background application cleanup instruction by clicking the specified cleanup icon.

2-2. Obtain a background application of the electronic device and current running parameters according to the background application cleanup instruction.

2-3. Calculate the cleanable rate of each background application by using the trained Bayesian model and the current operating parameters.

For example, the foregoing steps 2-3 may specifically include:

Determining a current feature value according to current operating parameters;

The current feature value is input into the third preset formula for calculation, and the cleanable rate is obtained. The third preset formula is:

Where 1 ≤ _k ≤ m, q _k is the current eigenvalue.

In this embodiment, similar to the training process, the current sampling date type, the sampling period, the preset power range, the preset duration range, the foreground application, the previous switching application, the charging connection state, and the network may be obtained according to the current operating parameters. The connection state and the 9 feature items of the background application that are currently to be predicted, then m is 9, and the feature values q ₁ , q ₂ ... q ₉ corresponding to the feature items are obtained, and then the formula is used:

P(j2|q ₁ ,q ₂ ...q ₉ )=P(j2)P(q ₁ |j2)P(q ₂ |j2)...P(q ₉ |j2) is calculated on the premise that the current eigenvalue occurs The probability of occurrence of j2 (that is, the background application that currently needs to be predicted will not switch to the foreground in a short time), as the cleanable rate.

2-4. The background application is closed according to the cleanup rate.

For example, the foregoing steps 2-4 may specifically include:

Select a background application whose cleaning rate is not less than the preset threshold as the target application, or select the preset number of background applications with the highest cleanup rate as the target application;

Close the target app.

In this embodiment, the preset threshold and the preset number can be manually set. For example, the preset threshold may be 0.5, and the preset number may be 4, that is, when the calculated P(j2|q ₁ When q ₂ ... q _m ) is greater than 0.5, it can be considered that the background application i will not switch to the foreground in a short time, and thus can be cleaned up as a cleanup object.

It can be seen from the above that the application control method provided in this embodiment is applied to an electronic device, and obtains an operation parameter of the electronic device at each sampling time point in a historical period, where the operation parameter includes a foreground application, a remaining power, a duration of a bright screen, and a charging connection state and a network connection state, and generating a training sample according to the sampling time point and the running parameter, and then training the preset Bayesian model by using the training sample, and based on the trained Bayesian model, the electronic device The background application in the middle is controlled, so that the background application that needs to be cleaned can be selected according to the usage of the previous application. The method is simple, the flexibility is high, the system resources are saved, and the user experience is good.

In this embodiment, the application control device is specifically integrated into the electronic device as an example for detailed description.

Referring to FIG. 2 and FIG. 3, an application control method may be as follows:

201. The electronic device acquires an operation parameter of each sampling time point in a historical period, where the operating parameter includes a foreground application, a remaining power, a screened duration, a charging connection state, and a network connection state.

For example, the historical time period may be the past month, and the sampling time point may be every minute in the past month. The running parameter may be extracted from a database, and the database may store the usage record of the application in the electronic device in the past month, the on-off record of the screen, the charging record, and the WiFi connection record, as shown in Table 1-4 below. Based on these records, the operating parameters for each sampling time point can be extracted.

Application name	Open the timestamp of this app
Com.tencent.mobileqq	1457550655465
Com.android.settings	1457605107522
...	...

Table 1: Application Usage Record

Screen status change	Timestamp
Bright->off	1457605131575
Off->light	1457605151786
...	...

Table 2: Bright screen off screen recording

State of charge change	Electricity	Timestamp
Enter charging	twenty three%	1457605131510
Out of charge	80%	1457605151786
...		...

Table 3: Charging record

Wifi status change	SSID	BSSID	Timestamp
Connect wifi	...	...	1457605111510
Disconnect wifi	...	...	1457605131286
...			...

Table 4: Wifi Record

202. The electronic device determines, according to the sampling time point, a sampling date type and a sampling period, and determines a preset power amount range to which the remaining power belongs, and determines a preset duration range to which the bright screen duration belongs.

For example, if the sampling time point is 10:55 on October 17, 2012, it can be divided into 48 time periods per day, then the day is Wednesday, the sampling date type is working day, and the sampling time is the 11th time period. If the remaining power is 80%, the preset power range can be 70%-100% of the corresponding high power. If the duration of the bright screen is 3 min, the preset duration may be a short duration corresponding to 0-5 min.

203. The electronic device acquires a target prediction application, and determines a previous switching application and a next switching application of the foreground application from the operating parameters according to the sampling time point.

For example, the target prediction application may be the ten most recent applications APP1, APP2...APP10. All foreground applications can be sorted according to the sampling time point. For any three adjacent different foreground applications after sorting, the front foreground application can be used as the last switching application of the intermediate foreground application, and the latter foreground application can be used as the middle. The next switching application of the foreground application, for example, for a sampling time point, the foreground application may be APP10, the last switching application may be APP1, and the next switching application may be APP5.

204. The electronic device calculates a difference between a sampling time point of the next switching application and a sampling time point of the foreground application, and determines whether the target prediction application is the next switching application, and whether the difference does not exceed a preset. The duration, if yes, determines the predicted value of the target prediction application as the first preset value, and if not, determines the predicted value of the target prediction application as the second preset value.

For example, the interval between APP1 and APP10 may be T1, the first preset value may be 1, the second preset value may be 0, and the preset duration may be 10 min. The prediction application is APP5, and if T1 ≤ 10, the predicted value of the target prediction application can be set to 1, otherwise, it is set to 0.

205. The electronic device separately obtains the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, and the feature value corresponding to the target prediction application, and according to the The eigenvalues and predicted values generate training samples.

For example, the composition of the training sample can be seen in FIG. 4, and the correspondence between the feature values and the feature items in the training sample can be seen in Table 5 below.

table 5

206. The electronic device inputs each training sample into a preset Bayesian model to train the Bayesian model.

For example, the eigenvalues may include (q ₁ , q ₂ ... q _m ), and the predicted value may include j1 and j2.

And inputting the predicted value into the first preset formula to obtain a probability of the corresponding predicted value, where the first preset formula is:

Where N(j1) represents the number of occurrences of event j1, N(j2) represents the number of occurrences of event j2, and P(j2) represents the probability of occurrence of event j2;

And inputting the feature value and the predicted value into the second preset formula to obtain a probability of the corresponding feature value and the predicted value, where the second preset formula is:

Where 1 ≤ _i ≤ m, P(q _i | j2) represents the probability that the event q _i occurs on the premise that the event j2 occurs, and N(q _i , j2) represents the number of times the events q _i and j2 occur simultaneously.

In this embodiment, the Bayesian model can be:

Where m is the number of feature terms, q ₁ , q ₂ ... q _m are the eigenvalues of the a priori condition, q _i is the eigenvalue corresponding to the i-th feature item, and J is the predicted value of the target prediction application. To simplify the calculation, assuming that q ₁ , q ₂ ... q _m are independent of each other, then

Thus the Naive Bayes classifier model is obtained:

J _MAX = arg max P(J|q ₁ , q ₂ ... q _m )=arg maxP(q ₁ |J)P(q ₂ |J)...P(q _m |J), where J can represent j1 or j2 The probability value of each feature item is the statistical probability of the number of occurrences, that is, the above formula:

Where j1 is the first preset value and j2 is the second preset value. It is easy to know that the process of training the Bayesian model is the process of probability and statistics, that is, after training the Bayesian model, the probability values of different eigenvalues in each feature item can be obtained, such as P(q ₁ ), P( q ₁ |j2).

207. The electronic device acquires a background application cleanup instruction.

For example, when it is detected that the memory usage reaches a certain limit, or the power is insufficient, or the running speed is too slow, the electronic device can automatically generate the background application cleaning instruction.

208. The electronic device acquires the background application and the current running parameter according to the background application cleanup instruction.

209. The electronic device calculates a cleanable rate of each background application by using the trained Bayesian model and the current operating parameters.

For example, the above step 209 may further include:

Determining a current feature value according to current operating parameters;

The current feature value is input into the third preset formula for calculation, and the cleanable rate is obtained. The third preset formula is:

Where 1 ≤ _k ≤ m, q _k is the current eigenvalue.

In this embodiment, similar to the training process, the current sampling date type, the sampling period, the preset power range, the preset duration range, the foreground application, the previous switching application, the charging connection state, and the network may be obtained according to the current operating parameters. The connection state and the 9 feature items of the background application that are currently to be predicted, then m is 9, and the feature values q ₁ , q ₂ ... q ₉ corresponding to the feature items are obtained, and then the formula is used:

P(j2|q ₁ ,q ₂ ...q ₉ )=P(j2)P(q ₁ |j2)P(q ₂ |j2)...P(q ₉ |j2) is calculated on the premise that the current eigenvalue occurs The probability of occurrence of j2 (that is, the background application that currently needs to be predicted will not switch to the foreground in a short time), as the cleanable rate.

210. The electronic device selects a background application whose cleaning rate is not less than a preset threshold as the target application, or selects a preset number of background applications with the highest cleanup rate as the target application, and closes the target application.

For example, the preset threshold may be 0.5, and the preset number may be 4, that is, when the calculated P(j2|q ₁ , q ₂ ... q _m ) is greater than 0.5, the background application i may be considered as a short time. It will not switch to the foreground and can be cleaned up as a cleanup object.

It can be seen from the above that the application control method provided in this embodiment, wherein the electronic device can acquire the running parameters of each sampling time point in the historical time period, the operating parameters include the foreground application, the remaining power, the screen duration, the charging connection state, and the network. a connection state, and then determining a sampling date type and a sampling period according to the sampling time point, determining a preset power amount range to which the remaining power belongs, and determining a preset duration range to which the bright screen duration belongs, and then acquiring a target prediction application And determining, according to the sampling time point, the previous switching application and the next switching application of the foreground application from the operating parameter, and then calculating a difference between the sampling time point of the next switching application and the sampling time point of the foreground application. And determining whether the target prediction application is the next handover application, and whether the difference does not exceed a preset duration, and if yes, determining the predicted value of the target prediction application as a first preset value, and if not, The predicted value of the target prediction application is determined as a second preset value, and then the previous handover is respectively obtained. a foreground sample, a sampling date type, a sampling period, a preset power range, a preset duration range, a charging connection state, a network connection state, and a feature value corresponding to the target prediction application, and generating a training sample according to the feature value and the predicted value, and then Each training sample is input into a preset Bayesian model to train the Bayesian model, and then a background application cleanup instruction is obtained, and the background application and the current running parameter are obtained according to the background application cleanup instruction. Then, using the trained Bayesian model and the current running parameters to calculate the cleanable rate of each background application, and selecting a background application whose cleaning rate is not less than a preset threshold as the target application, or selecting the highest cleanable rate The preset number of background applications is used as the target application, and then the target application is closed, so that the background application that needs to be cleaned can be selected according to the usage of the previous application, the method is simple, the flexibility is high, the system resources are saved, and the user experience is saved. Feeling good.

According to the method described in the foregoing embodiments, the present embodiment is further described from the perspective of an application control device, which may be implemented as an independent entity or integrated in an electronic device, such as a terminal. The terminal can include a mobile phone, a tablet computer, a personal computer, and the like.

An embodiment of the present application provides an application control apparatus, which is applied to an electronic device, and includes:

An acquiring module, configured to acquire an operating parameter of the electronic device at each sampling time point in a historical period, where the operating parameter includes a foreground application, a remaining power, a screened duration, a charging connection state, and a network connection state;

a generating module, configured to generate a training sample according to the sampling time point and the running parameter;

a training module, configured to train the preset Bayesian model by using the training sample;

And a control module, configured to control a background application in the electronic device based on the trained Bayesian model.

In some embodiments, the generating module comprises:

a first determining submodule, configured to determine a sampling date type and a sampling period according to the sampling time point;

a second determining sub-module, configured to determine a preset power range to which the remaining power belongs, and determine a preset duration range to which the bright time duration belongs;

And generating a submodule, configured to generate a training sample according to the sampling date type, a sampling period, a preset power range, a preset duration range, a foreground application, a charging connection state, and a network connection state.

In some embodiments, the generating submodule comprises:

An obtaining unit for acquiring a target prediction application;

a first determining unit, configured to determine, according to the sampling time point, a previous switching application and a next switching application of the foreground application from the operating parameters;

a second determining unit, configured to determine a predicted value of the target prediction application according to the sampling time point, a next switching application, and a foreground application;

a generating unit, configured to generate a training sample according to the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, the target prediction application, and the predicted value .

In some embodiments, the second determining unit is specifically configured to:

Calculating a difference between a sampling time point of the next switching application and a sampling time point of the foreground application;

Determining whether the target prediction application is the next handover application, and whether the difference does not exceed a preset duration;

If yes, determining the predicted value of the target prediction application as a first preset value;

If not, the predicted value of the target prediction application is determined as a second preset value.

In some embodiments, the generating unit is specifically configured to:

Obtaining, respectively, the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, and the feature value corresponding to the target prediction application;

A training sample is generated based on the feature value and the predicted value.

In some embodiments, the feature values include (q ₁ , q ₂ . . . q _m ), the predicted values include j1 and j2, and the training module is specifically configured to:

And inputting the predicted value into the first preset formula to obtain a probability of the corresponding predicted value, where the first preset formula is:

Where N(j1) represents the number of occurrences of event j1, N(j2) represents the number of occurrences of event j2, and P(j2) represents the probability of occurrence of event j2;

And inputting the feature value and the predicted value into the second preset formula to obtain a probability of the corresponding feature value and the predicted value, where the second preset formula is:

Where 1 ≤ _i ≤ m, P(q _i | j2) represents the probability that the event q _i occurs on the premise that the event j2 occurs, and N(q _i , j2) represents the number of times the events q _i and j2 occur simultaneously.

In some embodiments, the control module is specifically configured to:

Get background application cleanup instructions;

Obtaining, according to the background application cleanup instruction, a background application of the electronic device, and current running parameters;

Calculate the cleanable rate of each background application by using the trained Bayesian model and the current operating parameters;

The background application is closed according to the cleanup rate.

In some embodiments, the control module is specifically configured to:

Determining a current feature value according to current operating parameters;

The current feature value is input into the third preset formula for calculation to obtain a cleanable rate, and the third preset formula is:

Where 1 ≤ _k ≤ m, q _k is the current eigenvalue.

In some embodiments, the control module is specifically configured to:

Select a background application whose cleaning rate is not less than the preset threshold as the target application, or select the preset number of background applications with the highest cleanup rate as the target application;

Close the target app.

Referring to FIG. 5, FIG. 5 specifically describes an application control device provided by an embodiment of the present application, which is applied to an electronic device, which may include: an obtaining module 10, a generating module 20, a training module 30, and a control module 40, where:

(1) Acquisition module 10

The obtaining module 10 is configured to acquire an operating parameter of the electronic device at each sampling time point in the historical period, where the running parameter includes a foreground application, a remaining power, a screened duration, a charging connection state, and a network connection state.

In this embodiment, the historical period may be manually set, such as the previous month or the first two months. The sampling time point mainly refers to the sampling frequency, for example, it can be sampled every minute or every two minutes, and can be expressed in the form of x minutes x x x x x x. Both the charging connection state and the network connection state may include both connected and unconnected situations.

In the actual application process, the running parameter may be acquired in real time. For example, the acquiring module 10 obtains the corresponding data collecting operation at the sampling time point, or may be acquired at one time. For example, the electronic device may record the historical time period in the local database in advance. Each time the screen change data, the charging state change data, the network state change data, and the application open data are performed, the obtaining module 10 can extract the running parameters of each sampling time point at a time according to the sampling frequency.

(2) Generation module 20

The generating module 20 is configured to generate a training sample according to the sampling time point and the running parameter.

For example, referring to FIG. 6, the generating module 20 may specifically include a first determining submodule 21, a second determining submodule 22, and a generating submodule 22, where:

The first determining sub-module 21 is configured to determine a sampling date type and a sampling period according to the sampling time point.

In this embodiment, the sampling date type is divided into weekly, which may include weekdays and weekends. The sampling period is divided into daily, which can divide the day into 48 time periods.

The second determining sub-module 22 is configured to determine a preset power range to which the remaining power belongs, and determine a preset duration range to which the bright screen duration belongs.

In this embodiment, the preset power range and the preset duration range may be manually set, and the preset power range may include three ranges indicating high power, medium power, and low power, for example, the high power may be 70%. -100%, the medium power can be 40%-70%, the low battery can be 0-40%, etc., the preset duration range can include three intervals indicating short, medium and long, for example, the length can be more than 10min, It can be 5-10min in length and 0-5min in short.

The generating submodule 23 is configured to generate a training sample according to the sampling date type, the sampling period, the preset power range, the preset duration range, the foreground application, the charging connection state, and the network connection state.

For example, referring to FIG. 7, the generating sub-module 23 may specifically include an obtaining unit 231, a first determining unit 232, a second determining unit 233, and a generating unit 234, where:

The obtaining unit 231 is configured to acquire a target prediction application.

In this embodiment, the target prediction application may be all applications installed in the electronic device, or may be partial applications. When it is a partial application, it may be several applications with the highest frequency of occurrence in the near future, and the specific quantity may be according to actual needs. And set.

The first determining unit 232 is configured to determine, according to the sampling time point, the last switching application and the next switching application of the foreground application from the operating parameters.

In this embodiment, since the foreground application obtained by each sampling in the historical period is known, the foreground application obtained before the sampling time point can be regarded as the current foreground application for the foreground application obtained by any sampling. For the previous switching application, the different foreground applications obtained after the sampling time point can be regarded as the next switching application of the current foreground application. Generally, the first determining unit 232 can take different foreground applications that are closest to the current sampling time point. The last switch application and the next switch application. In the actual operation process, all foreground applications may be sorted according to the sampling time point. For any three adjacent different foreground applications after sorting, the front foreground application may serve as the previous switching application of the intermediate foreground application, followed by The foreground application can serve as the next switching application for the intermediate foreground application.

The second determining unit 233 is configured to determine a predicted value of the target prediction application according to the sampling time point, the next switching application, and the foreground application.

In this embodiment, the predicted value may be an artificially set value, such as 0 and 1, where 0 indicates that the target prediction application will not switch to the foreground use in a short time, and 1 may indicate that the target prediction application will be in the Switch to the foreground for a short time. Since all the foreground applications collected during the historical period are known, the predicted value of the target prediction application can be determined according to the known foreground application and the sampling time point thereof. At this time, the second determining unit 233 can further use to:

Calculating a difference between a sampling time point of the next switching application and a sampling time point of the foreground application;

Determining whether the target prediction application is the next handover application, and whether the difference does not exceed a preset duration;

If yes, determining the predicted value of the target prediction application as the first preset value;

If not, the predicted value of the target prediction application is determined as a second preset value.

In this embodiment, the preset duration, the first preset value, and the second preset value may be manually set, and the preset duration is mainly used to define a length of time, which may be 10 minutes, and the first preset value may be Yes 1, the second preset value can be 0. For each sampling, when the target prediction application that needs to be predicted is the next switching application, the second determining unit 233 needs to further analyze the length of time taken to switch from the current application to the next switching application, only when the interval duration is in advance. The predicted value of the target prediction application can be set to 1 when the duration is within, otherwise all are set to 0.

The generating unit 234 is configured to generate a training sample according to the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, the target prediction application, and the predicted value. .

In this embodiment, in order to analyze the user behavior from multiple dimensions, so that the trained machine learning model is more anthropomorphic, each training sample may be composed of a plurality of known feature items and data of the tag items, the known The feature item may include the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection status, and the network connection status, etc., and the label item is mainly.

For example, the generating unit 234 can be specifically configured to:

Obtaining, respectively, the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, and the feature value corresponding to the target prediction application;

A training sample is generated based on the feature value and the predicted value.

In this embodiment, since the computer program generally runs in the form of characters, the feature value can be mainly expressed in the form of Arabic numerals or letters, such as 1-10, and each feature item can also be expressed in the form of letters, such as the foreground. The application is H, the sampling date type is B, and so on. When generating the training sample, the generating unit 234 may directly use the feature value of the feature item as a prior condition, and use the predicted value of each target prediction application as a posterior result to generate the training sample.

It is easy to understand that the feature values corresponding to each feature item may be preset, and the feature values of different feature items may be the same or different, for example, the feature values of the foreground application and the sampling period may include 0 to 10, However, each number refers to a different meaning in different feature items. For example, for the foreground application, 0 can refer to the US group. For the sampling period, 0 can refer to the period from 0:00 to 1:00.

(3) Training module 30

The training module 30 is configured to train the preset Bayesian model with the training sample.

For example, the feature value may include (q ₁ , q ₂ ... q _m ), and the predicted value may include j1 and j2. In this case, the training module 30 may specifically be used to:

And inputting the predicted value into the first preset formula to obtain a probability of the corresponding predicted value, where the first preset formula is:

Where N(j1) represents the number of occurrences of event j1, N(j2) represents the number of occurrences of event j2, and P(j2) represents the probability of occurrence of event j2;

And inputting the feature value and the predicted value into the second preset formula to obtain a probability of the corresponding feature value and the predicted value, where the second preset formula is:

Where 1 ≤ _i ≤ m, P(q _i | j2) represents the probability that the event q _i occurs on the premise that the event j2 occurs, and N(q _i , j2) represents the number of times the events q _i and j2 occur simultaneously.

In this embodiment, the Bayesian model can be:

Where m is the number of feature terms, q ₁ , q ₂ ... q _m are the eigenvalues of the a priori condition, q _i is the eigenvalue corresponding to the i-th feature item, and J is the predicted value of the target prediction application. To simplify the calculation, assuming that q ₂ , q ₂ ... q _m are independent of each other, then

Thus the Naive Bayes classifier model is obtained:

J _MAX = arg max P(J|q ₁ , q ₂ ... q _m )=arg maxP(q ₁ |J)P(q ₂ |J)...P(q _m |J), where J can represent j1 or j2 The probability value of each feature item is the statistical probability of the number of occurrences, that is, the above formula:

Where j1 is the first preset value and j2 is the second preset value. It is easy to know that the process of training the Bayesian model is the process of probability and statistics, that is, after training the Bayesian model, the probability values of different eigenvalues in each feature item can be obtained, such as P(q ₁ ), P( q ₁ |j2).

(4) Control module 40

The control module 40 is configured to control the background application in the electronic device based on the trained Bayesian model.

For example, the control module 40 can be specifically configured to:

2-1. Obtain the background application cleanup command.

In this embodiment, the background application cleanup instruction may be automatically generated by the electronic device, for example, when the memory occupancy reaches a certain limit, or the power is insufficient, or the running speed is too slow, the background application cleanup instruction is generated, and of course, the background application is cleaned up. The instructions may also be generated manually by the user. For example, the user may generate the background application cleanup instruction by clicking the specified cleanup icon.

2-2. Obtain a background application of the electronic device and current running parameters according to the background application cleanup instruction.

2-3. Calculate the cleanable rate of each background application by using the trained Bayesian model and the current operating parameters.

For example, the foregoing control module 40 may specifically be used to:

Determining a current feature value according to current operating parameters;

The current feature value is input into the third preset formula for calculation, and the cleanable rate is obtained. The third preset formula is:

Where 1 ≤ _k ≤ m, q _k is the current eigenvalue.

In this embodiment, similar to the training process, the current sampling date type, the sampling period, the preset power range, the preset duration range, the foreground application, the previous switching application, the charging connection state, and the network may be obtained according to the current operating parameters. The connection state and the 9 feature items of the background application that are currently to be predicted, then m is 9, and the feature values q ₁ , q ₂ ... q ₉ corresponding to the feature items are obtained, and then the formula is used:

P(j2|q ₁ ,q ₂ ...q ₉ )=P(j2)P(q ₁ |j2)P(q ₂ |j2)...P(q ₉ |j2) is calculated on the premise that the current eigenvalue occurs The probability of occurrence of j2 (that is, the background application that currently needs to be predicted will not switch to the foreground in a short time), as the cleanable rate.

2-4. The background application is closed according to the cleanup rate.

For example, the control module 40 can further be used to:

Select a background application whose cleaning rate is not less than the preset threshold as the target application, or select the preset number of background applications with the highest cleanup rate as the target application;

Close the target app.

In this embodiment, the preset threshold and the preset number can be manually set. For example, the preset threshold may be 0.5, and the preset number may be 4, that is, when the calculated P(j2|q ₁ When q ₂ ... q _m ) is greater than 0.5, it can be considered that the background application i will not switch to the foreground in a short time, and thus can be cleaned up as a cleanup object.

In the specific implementation, the foregoing units may be implemented as a separate entity, or may be implemented in any combination, and may be implemented as the same or a plurality of entities. For the specific implementation of the foregoing, refer to the foregoing method embodiments, and details are not described herein.

It can be seen from the above that the application control device provided in this embodiment is applied to an electronic device, and the operation parameter of the electronic device is obtained by the acquisition module 10 at each sampling time point in the historical time period, and the operation parameter includes the foreground application, the remaining power, and the light is turned on. The screen duration, the charging connection state, and the network connection state. Then, the generating module 20 generates a training sample according to the sampling time point and the operating parameter, and the training module 30 uses the training sample to train the preset Bayesian model, and the control module 40 is based on The trained Bayesian model controls the background application in the electronic device, so that the background application that needs to be cleaned can be selected according to the usage of the previous application, the method is simple, the flexibility is high, and the system resources are saved, and the user Experience is good.

In addition, the embodiment of the present application further provides an electronic device, which may be a device such as a smart phone or a tablet computer. As shown in FIG. 8, the electronic device 500 includes a processor 501, a memory 502, a display screen 503, and a control circuit 504. The processor 501 is electrically connected to the memory 502, the display screen 503, and the control circuit 504, respectively.

The processor 501 is a control center of the electronic device 500, and connects various parts of the entire electronic device using various interfaces and lines, executes the electronic by running or loading an application stored in the memory 502, and calling data stored in the memory 502. The various functions and processing data of the device enable overall monitoring of the electronic device.

In this embodiment, the processor 501 in the electronic device 500 loads the instructions corresponding to the process of one or more applications into the memory 502 according to the following steps, and is stored and stored in the memory 502 by the processor 501. In the application, thus implementing various functions:

Obtaining an operating parameter of the electronic device at each sampling time point in the historical period, where the operating parameter includes a foreground application, a remaining power, a screened duration, a charging connection state, and a network connection state;

Generating a training sample according to the sampling time point and the running parameter;

Using the training sample to train the preset Bayesian model;

The background application in the electronic device is controlled based on the trained Bayesian model.

In some embodiments, the processor can be used to perform the following steps when generating training samples based on the sampling time points and operational parameters:

Determining a sampling date type and a sampling period according to the sampling time point;

Determining a preset power range to which the remaining power belongs, and determining a preset duration range to which the bright screen duration belongs;

The training samples are generated according to the sampling date type, the sampling period, the preset power range, the preset duration range, the foreground application, the charging connection state, and the network connection state.

In some embodiments, the processor can be used to perform the following steps when generating training samples according to the sampling date type, sampling period, preset power range, preset duration range, foreground application, charging connection status, and network connection status. :

Obtain a target prediction application;

Determining, according to the sampling time point, a previous switching application and a next switching application of the foreground application from the operating parameters;

Determining a predicted value of the target prediction application according to the sampling time point, a next switching application, and a foreground application;

A training sample is generated according to the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, the target prediction application, and the predicted value.

In some embodiments, the processor is operable to perform the following steps when determining the predicted value of the target prediction application based on the sampling time point, the next switching application, and the foreground application:

Calculating a difference between a sampling time point of the next switching application and a sampling time point of the foreground application;

Determining whether the target prediction application is the next handover application, and whether the difference does not exceed a preset duration;

If yes, determining the predicted value of the target prediction application as a first preset value;

If not, the predicted value of the target prediction application is determined as a second preset value.

In some embodiments, generating according to the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, the target prediction application, and the predicted value. When training a sample, the processor can be used to perform the following steps:

Obtaining, respectively, the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, and the feature value corresponding to the target prediction application;

A training sample is generated based on the feature value and the predicted value.

In some embodiments, the feature values include (q ₁ , q ₂ . . . q _m ), the predicted values include j1 and j2, when the preset Bayesian model is trained using the training samples, This processor can be used to perform the following steps:

And inputting the predicted value into the first preset formula to obtain a probability of the corresponding predicted value, where the first preset formula is:

Where N(j1) represents the number of occurrences of event j1, N(j2) represents the number of occurrences of event j2, and P(j2) represents the probability of occurrence of event j2;

And inputting the feature value and the predicted value into the second preset formula to obtain a probability of the corresponding feature value and the predicted value, where the second preset formula is:

Where 1 ≤ _i ≤ m, P(q _i | j2) represents the probability that the event q _i occurs on the premise that the event j2 occurs, and N(q _i , j2) represents the number of times the events q _i and j2 occur simultaneously.

In some embodiments, when the background application in the electronic device is controlled based on the trained Bayesian model, the processor can be used to perform the following steps:

Get background application cleanup instructions;

Obtaining, according to the background application cleanup instruction, a background application of the electronic device, and current running parameters;

Calculate the cleanable rate of each background application by using the trained Bayesian model and the current operating parameters;

The background application is closed according to the cleanup rate.

In some embodiments, the processor can be used to perform the following steps when calculating the cleanable rate of each background application using the trained Bayesian model and current operational parameters:

Determining a current feature value according to current operating parameters;

The current feature value is input into the third preset formula for calculation to obtain a cleanable rate, and the third preset formula is:

Where 1 ≤ _k ≤ m, q _k is the current eigenvalue.

In some embodiments, when the background application is closed according to the cleanup rate, the processor can be used to perform the following steps:

Select a background application whose cleaning rate is not less than the preset threshold as the target application, or select the preset number of background applications with the highest cleanup rate as the target application;

Close the target app.

Memory 502 can be used to store applications and data. The application stored in the memory 502 contains instructions executable in the processor. Applications can form various functional modules. The processor 501 executes various functional applications and data processing by running an application stored in the memory 502.

The display screen 503 can be used to display information entered by the user or information provided to the user as well as various graphical user interfaces of the terminal, which can be composed of images, text, icons, video, and any combination thereof.

The control circuit 504 is electrically connected to the display screen 503 for controlling the display screen 503 to display information.

In some embodiments, as shown in FIG. 8, the electronic device 500 further includes a radio frequency circuit 505, an input unit 506, an audio circuit 507, a sensor 508, and a power source 509. The processor 501 is electrically connected to the radio frequency circuit 505, the input unit 506, the audio circuit 507, the sensor 508, and the power source 509, respectively.

The radio frequency circuit 505 is used for transmitting and receiving radio frequency signals to establish wireless communication with network devices or other electronic devices through wireless communication, and to transmit and receive signals with network devices or other electronic devices.

The input unit 506 can be configured to receive input digits, character information, or user characteristic information (eg, fingerprints), and to generate keyboard, mouse, joystick, optical, or trackball signal inputs related to user settings and function controls. The input unit 506 can include a fingerprint identification module.

The audio circuit 507 can provide an audio interface between the user and the terminal through a speaker and a microphone.

Electronic device 500 may also include at least one type of sensor 508, such as a light sensor, motion sensor, and other sensors. Specifically, the light sensor may include an ambient light sensor and a proximity sensor, wherein the ambient light sensor may adjust the brightness of the display panel according to the brightness of the ambient light, and the proximity sensor may close the display panel and/or the backlight when the terminal moves to the ear. . As a kind of motion sensor, the gravity acceleration sensor can detect the magnitude of acceleration in all directions (usually three axes). When it is stationary, it can detect the magnitude and direction of gravity. It can be used to identify the gesture of the mobile phone (such as horizontal and vertical screen switching, related Game, magnetometer attitude calibration), vibration recognition related functions (such as pedometer, tapping), etc.; as for the terminal can also be configured with gyroscopes, barometers, hygrometers, thermometers, infrared sensors and other sensors, no longer Narration.

Power source 509 is used to power various components of electronic device 500. In some embodiments, the power supply 509 can be logically coupled to the processor 501 through a power management system to enable functions such as managing charging, discharging, and power management through the power management system.

Although not shown in FIG. 8, the electronic device 500 may further include a camera, a Bluetooth module, and the like, and details are not described herein again.

It will be understood by those skilled in the art that all or part of the steps of the various methods in the above embodiments may be completed by instructions or controlled by related hardware, which may be stored in a computer readable storage medium. And loaded and executed by the processor. To this end, an embodiment of the present invention provides a storage medium in which a plurality of instructions are stored, which can be loaded by a processor to perform the steps in any of the application control methods provided by the embodiments of the present invention.

The storage medium may include: a read only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

The steps in the application control method provided by the embodiments of the present invention may be implemented by using the instructions stored in the storage medium. Therefore, any application control method provided by the embodiments of the present invention may be implemented. For the beneficial effects, see the previous embodiments in detail, and details are not described herein again.

For the specific implementation of the foregoing operations, refer to the foregoing embodiments, and details are not described herein again.

In the above, although the present application has been disclosed in the above preferred embodiments, the preferred embodiments are not intended to limit the application, and those skilled in the art can make various modifications without departing from the spirit and scope of the application. The invention is modified and retouched, and the scope of protection of the present application is determined by the scope defined by the claims.

Claims (20)

1. An application control method is applied to an electronic device, including:

   Obtaining an operating parameter of the electronic device at each sampling time point in the historical period, where the operating parameter includes a foreground application, a remaining power, a screened duration, a charging connection state, and a network connection state;

   Generating a training sample according to the sampling time point and the running parameter;

   Training the preset Bayesian model with the training sample;

   The background application in the electronic device is controlled based on the trained Bayesian model.
2. The application control method according to claim 1, wherein the generating the training sample according to the sampling time point and the operating parameter comprises:

   Determining a sampling date type and a sampling period according to the sampling time point;

   Determining a preset power range to which the remaining power belongs, and determining a preset duration range to which the bright screen duration belongs;

   The training samples are generated according to the sampling date type, the sampling period, the preset power range, the preset duration range, the foreground application, the charging connection state, and the network connection state.
3. The application control method according to claim 2, wherein the generating a training sample according to the sampling date type, a sampling period, a preset power range, a preset duration range, a foreground application, a charging connection state, and a network connection state, including :

   Obtain a target prediction application;

   Determining, according to the sampling time point, a previous switching application and a next switching application of the foreground application from the operating parameters;

   Determining a predicted value of the target prediction application according to the sampling time point, a next switching application, and a foreground application;

   A training sample is generated according to the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, the target prediction application, and the predicted value.
4. The application control method according to claim 3, wherein the determining the predicted value of the target prediction application according to the sampling time point, the next switching application, and the foreground application comprises:

   Calculating a difference between a sampling time point of the next switching application and a sampling time point of the foreground application;

   Determining whether the target prediction application is the next handover application, and whether the difference does not exceed a preset duration;

   If yes, determining the predicted value of the target prediction application as a first preset value;

   If not, the predicted value of the target prediction application is determined as a second preset value.
5. The application control method according to claim 3, wherein the according to the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, and the network connection state , target prediction applications, and predictive value generation training samples, including:

   Obtaining, respectively, the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, and the feature value corresponding to the target prediction application;

   A training sample is generated based on the feature value and the predicted value.
6. The application control method according to claim 5, wherein said feature value comprises (q ₁ , q ₂ ... q _m ), said predicted value comprises j1 and j2, said using said training sample for a preset shell The Yesi model is trained to include:

   And inputting the predicted value into the first preset formula to obtain a probability of the corresponding predicted value, where the first preset formula is:

   

   Where N(j1) represents the number of occurrences of event j1, N(j2) represents the number of occurrences of event j2, and P(j2) represents the probability of occurrence of event j2;

   And inputting the feature value and the predicted value into the second preset formula to obtain a probability of the corresponding feature value and the predicted value, where the second preset formula is:

   

   Where 1 ≤ _i ≤ m, P(q _i | j2) represents the probability that the event q _i occurs on the premise that the event j2 occurs, and N(q _i , j2) represents the number of times the events q _i and j2 occur simultaneously.
7. The application control method according to claim 6, wherein the controlling the background application in the electronic device based on the trained Bayesian model comprises:

   Get background application cleanup instructions;

   Obtaining, according to the background application cleanup instruction, a background application of the electronic device, and current running parameters;

   Calculate the cleanable rate of each background application by using the trained Bayesian model and the current operating parameters;

   The background application is closed according to the cleanup rate.
8. The application control method according to claim 7, wherein the calculating the cleanable rate of each background application by using the trained Bayesian model and the current running parameters comprises:

   Determining a current feature value according to current operating parameters;

   The current feature value is input into the third preset formula for calculation to obtain a cleanable rate, and the third preset formula is:

   

   Where 1 ≤ _k ≤ m, q _k is the current eigenvalue.
9. The application control method according to claim 7, wherein the closing the background application according to the cleanup rate comprises:

   Select a background application whose cleaning rate is not less than the preset threshold as the target application, or select the preset number of background applications with the highest cleanup rate as the target application;

   Close the target app.
10. An application control device is applied to an electronic device, including:

    An acquiring module, configured to acquire an operating parameter of the electronic device at each sampling time point in a historical period, where the operating parameter includes a foreground application, a remaining power, a screened duration, a charging connection state, and a network connection state;

    a generating module, configured to generate a training sample according to the sampling time point and the running parameter;

    a training module, configured to train the preset Bayesian model by using the training sample;

    And a control module, configured to control a background application in the electronic device based on the trained Bayesian model.
11. The application control device according to claim 10, wherein the generating module comprises:

    a first determining submodule, configured to determine a sampling date type and a sampling period according to the sampling time point;

    a second determining sub-module, configured to determine a preset power range to which the remaining power belongs, and determine a preset duration range to which the bright time duration belongs;

    And generating a submodule, configured to generate a training sample according to the sampling date type, a sampling period, a preset power range, a preset duration range, a foreground application, a charging connection state, and a network connection state.
12. The application control device according to claim 11, wherein the generating submodule comprises:

    An obtaining unit for acquiring a target prediction application;

    a first determining unit, configured to determine, according to the sampling time point, a previous switching application and a next switching application of the foreground application from the operating parameters;

    a second determining unit, configured to determine a predicted value of the target prediction application according to the sampling time point, a next switching application, and a foreground application;

    a generating unit, configured to generate a training sample according to the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, the target prediction application, and the predicted value .
13. The application control device according to claim 12, wherein the second determining unit is specifically configured to:

    Calculating a difference between a sampling time point of the next switching application and a sampling time point of the foreground application;

    Determining whether the target prediction application is the next handover application, and whether the difference does not exceed a preset duration;

    If yes, determining the predicted value of the target prediction application as a first preset value;

    If not, the predicted value of the target prediction application is determined as a second preset value.
14. The application control device according to claim 12, wherein the generating unit is specifically configured to:

    Obtaining, respectively, the previous switching application, the foreground application, the sampling date type, the sampling period, the preset power range, the preset duration range, the charging connection state, the network connection state, and the feature value corresponding to the target prediction application;

    A training sample is generated based on the feature value and the predicted value.
15. The application control device according to claim 14, wherein the feature value comprises (q ₁ , q ₂ ... q _m ), the predicted value comprises j1 and j2, and the training module is specifically configured to:

    And inputting the predicted value into the first preset formula to obtain a probability of the corresponding predicted value, where the first preset formula is:

    

    Where N(j1) represents the number of occurrences of event j1, N(j2) represents the number of occurrences of event j2, and P(j2) represents the probability of occurrence of event j2;

    And inputting the feature value and the predicted value into the second preset formula to obtain a probability of the corresponding feature value and the predicted value, where the second preset formula is:

    

    Where 1 ≤ _i ≤ m, P(q _i | j2) represents the probability that the event q _i occurs on the premise that the event j2 occurs, and N(q _i , j2) represents the number of times the events q _i and j2 occur simultaneously.
16. The application control device according to claim 15, wherein the control module is specifically configured to:

    Get background application cleanup instructions;

    Obtaining, according to the background application cleanup instruction, a background application of the electronic device, and current running parameters;

    Calculate the cleanable rate of each background application by using the trained Bayesian model and the current operating parameters;

    The background application is closed according to the cleanup rate.
17. The application control device according to claim 16, wherein the control module is specifically configured to:

    Determining a current feature value according to current operating parameters;

    The current feature value is input into the third preset formula for calculation to obtain a cleanable rate, and the third preset formula is:

    

    Where 1 ≤ _k ≤ m, q _k is the current eigenvalue.
18. The application control device according to claim 16, wherein the control module is specifically configured to:

    Select a background application whose cleaning rate is not less than the preset threshold as the target application, or select the preset number of background applications with the highest cleanup rate as the target application;

    Close the target app.
19. A computer readable storage medium, wherein the storage medium stores a plurality of instructions adapted to be loaded by a processor to perform the application control method of claim 1.
20. An electronic device comprising a processor and a memory, the processor being electrically connected to the memory, the memory for storing instructions and data, the processor being configured to execute the application control method according to claim 1 A step of.

Images