Application recommendation prediction method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention discloses an application recommendation prediction method and device, electronic equipment and a storage medium. The method comprises the steps of obtaining a knocking instruction and environment information; and starting a target application program based on the knock instruction and the environment information. The prediction method for application recommendation, provided by the embodiment of the invention, has relatively high prediction flexibility.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method, device, electronic device and storage medium for predicting application recommendations. Background Art

[0002] In the prior art, it is possible to quickly launch an application by tapping the back of an electronic device, that is, quickly open an application when the screen is on or off. However, this solution is limited by the correspondence between the tapping method and the application, that is, one tapping method can only be used to open a preset application, and it is not possible to flexibly open the application currently required by the user according to the state of the user or the electronic device. Summary of the invention

[0003] Embodiments of the present application provide a method, device, electronic device, and storage medium for predicting application recommendations.

[0004] In a first aspect, an embodiment of the present application provides a prediction method for application recommendation, which is applied to an electronic device, including: obtaining a tapping instruction and environmental information; and starting a target application based on the tapping instruction and the environmental information.

[0005] In some implementations, launching the target application based on the tapping instruction and the environmental information includes: inputting the tapping instruction and the environmental information into a target prediction model to obtain the target application.

[0006] In some implementations, the training method of the target prediction model includes: constructing a training data set, wherein the training data set includes multiple training sample pairs, each of the training sample pairs includes a tapping instruction sample, an environmental information sample, and a corresponding target application label; and training an initial prediction model according to the training data set to obtain a target prediction model.

[0007] In some implementations, the training method of the target prediction model also includes: obtaining a personalized data set, wherein the personalized data set includes at least one personalized sample pair, each of the personalized sample pairs includes the tapping instruction sample, the environmental information sample, and a corresponding personalized application label; and updating the target prediction model according to the personalized data set.

[0008] In some implementations, the method further includes: determining whether the target application is correct; and recommending at least one alternative application if the target application is incorrect.

[0009] In some implementations, the method further includes: determining a personalized application from the at least one candidate application, wherein the personalized application is used to update the target prediction model.

[0010] In some implementations, the tap instruction includes the location and / or number of times the electronic device was tapped.

[0011] In a second aspect, an embodiment of the present application provides a prediction device for application recommendation, including: an acquisition module for acquiring tapping instructions and environmental information; and a prediction module for launching a target application based on the tapping instructions and the environmental information.

[0012] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the operations in the method described in the first aspect when executing the computer program.

[0013] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the operations in the method described in the first aspect are implemented.

[0014] In summary, the embodiments of the present application disclose a prediction method, device, electronic device and storage medium for application recommendation. The method includes obtaining a tapping instruction and environmental information; and starting a target application based on the tapping instruction and the environmental information. The prediction method for application recommendation provided by the embodiments of the present application has high prediction flexibility. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of this specification, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0016] Figure 1 A schematic diagram showing a flow chart of a prediction method for application recommendation provided by some embodiments of the present application;

[0017] Figure 2 A schematic diagram of a process of training a target prediction model provided by some embodiments of the present application is shown;

[0018] Figure 3 A schematic diagram showing a flow chart of a prediction method for application recommendation provided by other embodiments of the present application;

[0019] Figure 4 An example diagram showing alternative application recommendations provided by some embodiments of the present application;

[0020] Figure 5 A schematic diagram showing the structure of a prediction device for application recommendation provided in some embodiments of the present application is shown;

[0021] Figure 6 A schematic diagram of the structure of an electronic device provided in some embodiments of the present application is shown. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in this application to clearly and completely describe the technical solutions in this application. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments in the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0023] Figure 1 A flowchart of a prediction method for application recommendation provided by some embodiments of the present application is shown. Although a logical order is shown in the flowchart, in some cases, the operations shown or described may be performed in an order different from that shown in the accompanying drawings. The prediction method can be applied to electronic devices, which may include but are not limited to smart phones, tablet computers, in-vehicle smart terminals and other devices. The specific process of the prediction method can be as follows:

[0024] S101, obtaining tapping instructions and environmental information.

[0025] Among them, the tapping instruction may include the position and / or number of times the electronic device is tapped. The position of the electronic device being tapped may be the front, back, top, bottom or side. Taking the back as an example, the back may further specifically include the upper part, middle part, lower part, etc. of the back. The tapping instruction may include tapping once or multiple times at the same position, or tapping at different positions respectively.

[0026] The tapping instruction can be obtained through sensors in the electronic device. In some embodiments, the electronic device can monitor whether there is signal fluctuation in the acceleration sensor and the gyroscope, and obtain the position and / or number of taps based on the acceleration sensor signal and the gyroscope signal. In some embodiments, the electronic device can obtain pressure size and distribution data through a pressure sensor, and determine the tapping position and / or number based on the pressure data. In other embodiments, the electronic device can also detect vibration waves through a vibration sensor, and determine the tapping position and / or number by analyzing the vibration waveform.

[0027] The environmental information may include at least one of time information, weather information, geographic location information or network connection status information.

[0028] In some embodiments, the time information may include a date and a time, where the date may include a specific year, month, day, Monday to Sunday, or may also include whether it is a holiday, and the time is a specific time point, which may only be accurate to a certain time period or number of hours. The time information may be obtained through a real-time clock built into an electronic device or from a time server or an operator base station, and this application does not limit this.

[0029] In some embodiments, weather information may include information reflecting weather conditions such as sunny, cloudy, rainy, windy, snowy, temperature, humidity, air quality, ultraviolet intensity, etc. Weather information may be obtained directly from a weather service provider or from a weather application.

[0030] In some embodiments, the geographic location information may include special locations and non-special locations. Among them, the special location may include home, company, station, shopping mall or other location marked by the user. The geographic location information can be obtained through a satellite positioning system, such as the Beidou satellite navigation system, the GPS positioning system, etc. The geographic location information can also be obtained through a signal base station, which is not limited in this application.

[0031] In some embodiments, the network connection information may include connection status information, network type information, or network quality information. Among them, the connection status information may include whether the Internet is connected or not connected. The network type information may include a data network connection or a WIFI connection. The network quality may include information such as network speed and bandwidth. The electronic device may query or monitor the network connection information of the local device through a network connection information query interface, or obtain the network connection information through other means, which is not limited in this application.

[0032] S102: Start a target application based on the tapping instruction and environmental information.

[0033] In some embodiments, starting the target application based on the tapping instruction and the environmental information may be to determine an application that may need to be opened based on the tapping instruction, and then determine the target application from the applications that may need to be opened in combination with the environmental information.

[0034] To this end, applications corresponding to different tapping instructions can be preset. For example, the tapping instruction is to tap once on the back, and the corresponding application may be a smart home control application. For example, the tapping instruction is to tap twice on the back, and the corresponding application may be a shopping application. For example, the tapping instruction is to tap once on the side, and the corresponding application may be a game application. For another example, the tapping instruction is to tap once on the top, and the corresponding application may be a social media application. In other implementations, applications corresponding to different tapping instructions may also be preset based on the manual settings of the user. For example, the tapping instruction is to tap once on the back, and the corresponding application may be a takeaway application, a transportation application, and an instant chat application set by the user.

[0035] The launch probability of different applications can be associated with different environmental information. For example, applications with high network requirements such as game applications or video applications have a significant correlation with WIFI connection. For another example, office applications have a significant correlation with the company's geographic location information, and smart home control applications have a significant correlation with the home's geographic location information. Therefore, the target application that needs to be launched can be determined by combining the tapping instruction with the environmental information.

[0036] In other embodiments, starting the target application based on the tapping instruction and the environmental information may be performed by inputting the tapping instruction and the environmental information into a target prediction model to obtain the target application.

[0037] Among them, the target prediction model is used to determine the target application based on the tapping instructions and environmental information. The target prediction model can be a classification model or a natural language model. For example, the target prediction model is a classification model. The input of the classification model can be feature data after feature digitization processing of the tapping instructions and environmental information, and the result of the model output can be the probability that each application is the target application, and the application with the highest probability can be used as the target application. For example, the target prediction model is a natural language model. The data input of the natural language model can be feature data obtained by feature embedding of the tapping instructions and environmental information, and the result of the model output can be text containing the name of the target application. The target prediction model can determine the target application that needs to be started based on the characteristics of the tapping instructions and environmental information.

[0038] The training of the target prediction model can be performed on the server. After the target prediction model is trained, the electronic device can obtain the target prediction model to perform the prediction reasoning process locally, or upload the data of the tapping instruction and environmental information to the server where the target prediction model is located for prediction reasoning, and then obtain the reasoning result.

[0039] Some embodiments of the present application provide a prediction method for application recommendation, including obtaining a tapping instruction and environmental information; and starting a target application based on the tapping instruction and environmental information. Some embodiments of the present application provide a prediction method with high prediction flexibility.

[0040] Next, how to obtain the target prediction model is described in detail.

[0041] Figure 2 A flow chart of a training target prediction model provided by some embodiments of the present application is shown. Although a logical order is shown in the flow chart, in some cases, the operations shown or described may be performed in an order different from that shown in the accompanying drawings. The specific process of training the target prediction model may be as follows:

[0042] S201: Build a training data set.

[0043] The training data set includes multiple training sample pairs, each training sample pair includes a tapping instruction sample, an environmental information sample and a corresponding target application label.

[0044] In some embodiments, the target prediction model is a classification model, and the tapping instruction sample can be binary feature data obtained by encoding the tapping instruction. For example, the tapping instruction is to tap once on the back, and the corresponding feature data can be "00010001"; the tapping instruction is to tap twice on the side, and the corresponding feature data can be "00100010". The encoding method can be to set different meanings for data at specific locations. For example, the first 4 bits are used to indicate the tapping position, and the last 4 bits indicate the number of tapping times. The encoding method can also be to encode all tapping methods in sequence. For example, the feature data of the first tapping method is "00000001", and the feature data of the second tapping method is "00000010".

[0045] The environmental information sample may be binary feature data obtained by encoding the environmental information. Similar to the tapping instruction sample, the encoding method may be to set different meanings for data at a specific location, for example, the first 9 digits encode the date, the next 5 digits represent the time, the next 3 digits encode Monday to Sunday, the next 5 digits encode the geographic location, and the last 4 digits encode the network connection information. "00001111100001011000010001" may represent the 31st day of this year, 1 a.m., Wednesday, home, and a stable WIFI connection.

[0046] The target application label may be label data obtained by encoding different applications. One-hot encoding may be used for encoding. For example, if there are 20 applications in total, the label data has 20 bits in total, each bit corresponding to one application. For example, the label corresponding to the third application is "000000000000000000100".

[0047] In some embodiments, the target prediction model is a natural language model, the tapping instruction sample may be feature data obtained by embedding features into text describing the tapping instruction, the environmental information sample may be feature data obtained by embedding features into text describing the environmental information, and the corresponding target application label may be the name of the target application.

[0048] Each tapping instruction sample and environmental information sample are concatenated to obtain a training sample, and the target application corresponding to the tapping instruction and the environmental information is used as its target application label to obtain multiple training sample pairs. The target application corresponding to the tapping instruction and the environmental information can be determined by a large number of experimentally collected data that conforms to the usage habits of ordinary users.

[0049] S202: Train the initial prediction model according to the training data set to obtain a target prediction model.

[0050] The initial prediction model may be a model obtained by pre-training on a large-scale data set. The initial prediction model is fine-tuned using the constructed training data set, wherein the fine-tuning method includes but is not limited to full parameter adjustment and partial parameter adjustment. Partial parameter adjustment may adopt a low-rank adapter (LoRA) or other variant adjustment methods. When the adjustment of the initial prediction model reaches the preset stop condition, the training is stopped to obtain the trained target prediction model.

[0051] In some embodiments, in order to enable the target prediction model to perform personalized predictions for different users, the training method of the target prediction model may further include:

[0052] S203: Obtain a personalized data set.

[0053] Among them, the personalized data set includes at least one personalized sample pair, and each personalized sample pair includes a tapping instruction sample, an environmental information sample, and a corresponding personalized application label. The personalized application label can be an application label determined based on a specific user's usage habits. For example, the target prediction model obtained after training in operation S202 tends to recommend travel applications at the subway station, and user A wants to open a music application at the subway station. In this case, the target application corresponding to the tapping instruction and environmental information in this scenario should be a music application. At this time, it is necessary to collect the tapping instruction, environmental information, and the application you want to open and use them as a personalized sample pair. The method of obtaining a tapping instruction sample based on the tapping instruction, obtaining an environmental information sample based on the environmental information, and obtaining a personalized label based on the application you want to open is similar to the training method in the previous text, and will not be repeated here.

[0054] S204: Update the target prediction model according to the personalized data set.

[0055] In some embodiments, the target prediction model is updated once each personalized sample pair is obtained. In other embodiments, an update may be performed after a certain number of personalized sample pairs are obtained. The update and fine-tuning of the target prediction model may be performed in an electronic device, or the personalized data set may be uploaded to a server for model update and fine-tuning.

[0056] The target prediction model training method provided in some embodiments of the present application can optimize and adjust the target prediction model according to the usage habits of different users, so that it is more in line with the user's personalized usage habits.

[0057] Figure 3 A schematic flow chart of a prediction method for application recommendation provided by other embodiments of the present application is shown. Although a logical order is shown in the flow chart, in some cases, the operations shown or described may be performed in an order different from that shown in the accompanying drawings. The specific process of the prediction method may be as follows:

[0058] S301, obtaining tapping instructions and environmental information.

[0059] S302: Start a target application based on the tapping instruction and environmental information.

[0060] S303: Determine whether the target application is correct.

[0061] In some embodiments, the electronic device may ask the user whether the currently opened target application is the application that the user needs to open, or ask the user whether other applications need to be opened, through a pop-up window or the like. Generally, information about whether the target application is correct may be obtained through "yes" and "no" buttons. For example, the electronic device asks "Is the opened application correct?" through a pop-up window. If the "yes" button is clicked, it indicates that the opened target application is correct; if the "no" button is clicked, it indicates that the opened target application is incorrect.

[0062] In other embodiments, the electronic device may monitor the time interval between the start of the target application and the next tapping instruction. If the next tapping instruction is received within a preset time interval after the target application is started last time, it is assumed that the previous target application is incorrect.

[0063] S304: When the target application is incorrect, recommend at least one alternative application.

[0064] The candidate application may be other application that may need to be opened except the target application.

[0065] In some embodiments, the target prediction model is a classification model, and the candidate applications can be determined according to the probability that each application output by the classification model is the target application. For example, the probabilities are sorted, and the candidate applications are determined starting from the application with the second highest probability according to the number of preset candidate applications.

[0066] In other embodiments, the target prediction model is a natural language model, and the candidate applications can be determined based on the application names in the text output by the natural language model. For example, the result of the model output can be "xx music, xx podcast, or xx radio". When the target application is "xx music", which is incorrect, "xx podcast" and "xx radio" are recommended as candidate applications.

[0067] In some embodiments, the prediction method may further include:

[0068] S305: Determine a personalized application from at least one candidate application.

[0069] The personalized application may be determined according to a user's selection from among the candidate applications. Figure 4 An example diagram showing alternative application recommendations provided by some embodiments of the present application is shown.

[0070] like Figure 4As shown, after starting the target application based on the tapping instruction and environmental information, the electronic device can ask "Do you need to open other applications?" through a pop-up window. If the "Yes" button is selected, it means that the opened target application is incorrect. If the "No" button is selected, it means that the opened target application is correct. If other applications need to be opened, the application to be opened can be determined through the pop-up window. Figure 4 As shown in , three candidate applications are recommended, and the user can choose from "Application 1", "Application 2", and "Application 3". In other embodiments, the pop-up window may also include a "None of the above applications is correct" or a similar button.

[0071] In some embodiments, in order to enable the target prediction model to perform personalized predictions for different users, a personalized application is used to update the target prediction model. As described above in operations S203-S204, a personalized data set can be obtained through the user's selection of the personalized application, and the target prediction model is optimized and updated based on the personalized data set.

[0072] Figure 5 The schematic diagram of the structure of the prediction device for application recommendation provided by some embodiments of the present application is shown. The prediction device 500 may specifically include:

[0073] The acquisition module 501 can be used to acquire tapping instructions and environmental information.

[0074] The prediction module 502 may be used to start a target application based on the tapping instruction and the environment information.

[0075] The prediction device 500 provided in some embodiments of the present application may firstly acquire the tapping instruction and the environment information by the acquisition module 501, and then the prediction module 502 may start the target application based on the tapping instruction and the environment information. The prediction device 500 provided in some embodiments of the present application has high prediction flexibility.

[0076] In some embodiments, the prediction module 502 may include a reasoning unit. The reasoning unit may be used to input the tapping instruction and the environment information into the target prediction model to obtain the target application.

[0077] In some embodiments, the prediction device 500 may further include a training module. The training module may be used to train the target prediction model. The training module may include a first data unit and a first training unit. The first data unit may be used to construct a training data set, wherein the training data set includes a plurality of training sample pairs, each training sample pair including a tapping instruction sample, an environmental information sample, and a corresponding target application label. The first training unit may be used to train the initial prediction model according to the training data set to obtain a target prediction model.

[0078] In some embodiments, the training module may further include a second data unit and a second training unit. The second data unit may be used to obtain a personalized data set, wherein the personalized data set includes at least one personalized sample pair, each personalized sample pair includes a tapping instruction sample, an environmental information sample, and a corresponding personalized application label. The second training unit may be used to update the target prediction model according to the personalized data set.

[0079] In some embodiments, the prediction device 500 may further include a judgment module and a recommendation module. The judgment module may be used to judge whether the target application is correct. The recommendation module may be used to recommend at least one alternative application if the target application is incorrect.

[0080] In some embodiments, the prediction device 500 may further include a determination module. The determination module may be used to determine a personalized application from at least one candidate application, wherein the personalized application is used to update the target prediction model.

[0081] In some embodiments, the tap instructions include the location and / or number of times the electronic device was tapped.

[0082] In addition, the present application also provides an electronic device, Figure 6 The following is a schematic diagram of the structure of an electronic device provided by some embodiments of the present application, specifically:

[0083] The electronic device 600 may include one or more processors 601 of processing cores, one or more computer-readable storage media memories 602 and other components. Those skilled in the art will appreciate that Figure 6 The structure of the electronic device 600 shown in the figure does not constitute a limitation on the electronic device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0084] The processor 601 is the control center of the electronic device 600. It uses various interfaces and lines to connect various parts of the entire electronic device 600. By running or executing software programs and / or modules stored in the memory 602, and calling data stored in the memory 602, it executes various functions of the electronic device 600 and processes data, thereby monitoring the electronic device 600 as a whole. Optionally, the processor 601 may include one or more processing cores; preferably, the processor 601 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, and the modem processor mainly processes wireless communications. It is understandable that the above-mentioned modem processor may not be integrated into the processor 601.

[0085] The memory 602 can be used to store software programs and modules. The processor 601 executes various functional applications and data processing by running the software programs and modules stored in the memory 602. The memory 602 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function, etc.; the data storage area may store data created according to the use of the electronic device 600, etc. In addition, the memory 602 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices. Accordingly, the memory 602 may also include a memory controller to provide the processor 601 with access to the memory 602.

[0086] Specifically in this embodiment, the processor 601 in the electronic device 600 will load the executable files corresponding to the processes of one or more applications into the memory 602 according to the following instructions, and the processor 601 will run the applications stored in the memory 602, thereby implementing the operations in any application recommendation prediction method provided in the embodiments of the present application.

[0087] The specific process of the electronic device 600 executing the prediction method for application recommendation can be referred to Figures 1 to 3 The description in , this application will not be repeated here.

[0088] Those skilled in the art will appreciate that all or part of the operations in the various methods of the above embodiments may be accomplished through instructions, or through instruction-controlled related hardware. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0089] To this end, the present application provides a computer-readable storage medium having a computer program stored thereon, and the computer program can be loaded by a processor to execute operations in any application-recommended prediction method provided in the present application.

[0090] The specific implementation of the above operations can be found in the previous embodiments, which will not be described in detail here.

[0091] The computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0092] Since the instructions stored in the computer-readable storage medium can execute the operations in any prediction method recommended by the application provided in the present application, the beneficial effects that can be achieved by any prediction method recommended by the application provided in the present application can be achieved. Please refer to the previous embodiments for details and will not be repeated here.

[0093] The above is a detailed introduction to the prediction method, device, electronic device and storage medium for application recommendation provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A prediction method for application recommendation, applied to electronic equipment, characterized in that: include: Get tapping instructions and environmental information; as well as A target application is started based on the tapping instruction and the environmental information.

2. The prediction method according to claim 1, characterized in that: The starting the target application based on the tapping instruction and the environmental information includes: The tapping instruction and the environmental information are input into a target prediction model to obtain the target application.

3. The prediction method according to claim 2, characterized in that: The training method of the target prediction model includes: Constructing a training data set, wherein the training data set includes a plurality of training sample pairs, each of the training sample pairs includes a tapping instruction sample, an environmental information sample, and a corresponding target application label; and The initial prediction model is trained according to the training data set to obtain a target prediction model.

4. The training method according to claim 3, characterized in that: The training method of the target prediction model also includes: Acquire a personalized data set, wherein the personalized data set includes at least one personalized sample pair, each of the personalized sample pairs includes the tapping instruction sample, the environmental information sample, and a corresponding personalized application label; and The target prediction model is updated according to the personalized data set.

5. The prediction method according to claim 2, characterized in that: Also includes: Determining whether the target application is correct; as well as In the event that the target application is incorrect, at least one alternative application is recommended.

6. The prediction method according to claim 5, characterized in that: Also includes: A personalized application is determined from the at least one candidate application, wherein the personalized application is used to update the target prediction model.

7. The prediction method according to claim 1, characterized in that: The tapping instruction includes the location and / or number of times the electronic device is tapped.

8. A prediction device for application recommendation, characterized in that: include: The acquisition module is used to obtain the tapping instructions and environmental information; as well as A prediction module is used to start a target application based on the tapping instruction and the environmental information.

9. An electronic device, characterized in that: The method comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the operations in the method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the operations in the method according to any one of claims 1 to 7 are implemented.

Images