Application recommendation method, electronic device and computer program product

Abstract

The invention provides an application recommendation method, electronic equipment and a computer program product, and relates to the technical field of application services. The method comprises the following steps: acquiring current scene data and state data of a target carrier; the scene data comprises at least one of current weather data, lane data and distance information between the target carrier and the service point; the state data comprises at least one of current carrier speed, carrier position, carrier temperature, carrier personnel data and hardware state; substituting the scene data and the state data into a prediction model corresponding to the target carrier for prediction to obtain a recommendation scheme comprising at least one recommendation application; the prediction model is constructed based on historical scene data, historical state data and historical application data of the target carrier. According to the technical scheme, the prediction model can be used, prediction is carried out in combination with the current scene data and the state data, targeted recommendation is carried out, the correlation between the recommendation scheme and the actual demand of the user is effectively improved, and therefore the use experience of the user is optimized.

Description

Technical Field

[0001] The present application relates to the field of application service technology, and in particular to an application recommendation method, electronic device, and computer program product. Background Art

[0002] In order to provide users with better services on the road, application recommendation functions are usually added to terminal devices such as in-vehicle systems to recommend various types of service applications and provide users with various types of services such as music, navigation, parking / charging guidance, etc., thereby improving the user's driving experience.

[0003] However, the current application recommendation function is usually performed by developers or operators before the release of the application, manually configuring the recommendation weights of each type of service application based on their own experience. As a result, the recommended application solutions are relatively fixed and inflexible, and personalized recommendations cannot be made based on the actual needs of users. As a result, the recommended application services are disconnected from the actual needs of users, and the probability of users using the application services is low, resulting in a poor user experience and an inability to provide effective services to users. Summary of the Invention

[0004] In view of this, the purpose of the embodiments of the present application is to provide an application recommendation method, an electronic device, and a computer program product to improve the problem of poor application service usage experience in the prior art.

[0005] To solve the above problems, in a first aspect, an embodiment of the present application provides an application recommendation method, the method comprising:

[0006] Acquiring current scene data and status data of the target carrier; wherein the scene data includes at least one of current weather data, lane data, and distance information between the target carrier and the service point; and the status data includes at least one of current carrier speed, carrier position, carrier temperature, carrier personnel data, and hardware status;

[0007] Substituting the scenario data and the status data into a prediction model corresponding to the target carrier for prediction, a recommendation solution including at least one recommended application is obtained; wherein the prediction model is constructed based on historical scenario data, historical status data and historical application data of the target carrier.

[0008] In the above implementation process, the historical scene data, historical status data and historical application data of the target carrier can be used to build a prediction model for predicting the usage of application services. When making application recommendations, the current scene data and status data of the target carrier can be obtained, and the current data can be used as input data in the input prediction model for prediction, and a recommendation plan containing one or more recommended applications can be obtained. Targeted application recommendations can be made based on the current actual situation and historical habits of the target carrier, effectively improving the correlation between the recommendation plan and the actual needs of the user, so as to increase the probability of the target carrier using the recommended application, thereby optimizing the user's usage experience of the application service.

[0009] Optionally, the prediction model is constructed in the following manner:

[0010] In conjunction with historical time nodes, training data and label data are constructed based on the historical scene data, historical status data, and historical application data of the target carrier obtained; wherein the historical scene data, the historical status data, and the historical application data are associated through the historical time nodes; the historical scene data includes at least one of historical weather data, historical lane data, and historical distance information between the target carrier and the service point; the historical status data includes at least one of historical carrier speed, historical carrier position, historical carrier temperature, historical carrier personnel data, and historical hardware status; and the historical application data includes historical application identity information and historical application usage information of historical applications selected by users;

[0011] Training an initial prediction model based on the training data, and verifying a prediction result of the initial prediction model based on the label data;

[0012] When the prediction accuracy of the prediction result is greater than or equal to a preset accuracy threshold, the prediction model is obtained.

[0013] In the above implementation process, the historical data of the target carrier, such as historical scene data, historical status data, and historical application data, are all associated with historical time nodes. Therefore, the historical time nodes and various historical data can be combined to construct corresponding training data and label data, and the training data is used to train the initial prediction model. The label data is used to verify the prediction results of the initial prediction model during the training process. When the prediction accuracy of the prediction result is greater than or equal to the preset accuracy threshold, it indicates that the current prediction accuracy of the initial prediction model is high and meets the usage requirements. The training can be stopped and the current initial prediction model can be used as the prediction model. The corresponding prediction model can be trained according to the historical scene conditions, historical status conditions, and historical usage of the application, and the prediction accuracy of the prediction model can be effectively improved by verifying the prediction results, further optimizing the correlation between the recommendation scheme determined by the prediction model and the actual demand situation.

[0014] Optionally, the method further includes:

[0015] Determining a plurality of different training weights according to different time periods corresponding to the training data; wherein the training weight is inversely proportional to the time distance between the time period and the current time node;

[0016] The prediction model is updated based on a preset update cycle and the training weights.

[0017] In the above implementation process, considering that the usage habits of users in the target carrier will change accordingly over time, when using the prediction model, multiple training weights of different sizes can be determined based on the training data of different time periods. The training weights corresponding to time periods close to the current time node are larger, and the training weights corresponding to time periods far from the current time node are smaller. In combination with the set update cycle and training weights, the prediction model is iteratively updated or retrained. The prediction model can be updated mainly based on recent training data, thereby effectively sensing changes in user usage habits and periodically updating the prediction model in real time based on changes in habits, thereby improving the real-time and effectiveness of the prediction model.

[0018] Optionally, the method further includes:

[0019] Get the recent historical data within the preset update period before the current time node;

[0020] The prediction model is updated based on the recent historical data and the update cycle.

[0021] In the above implementation, a corresponding update cycle can be set during the use of the prediction model for scheduled updates. During updates, recent historical data within the update cycle before the current time node is obtained and used as training data to iteratively update the current prediction model. The prediction model can be periodically updated using the recent historical data within each update cycle as training data, effectively improving the update efficiency and effectiveness of the prediction model.

[0022] Optionally, the step of substituting the scenario data and the state data into a prediction model corresponding to the target carrier to perform prediction and obtain a recommendation solution including at least one recommended application includes:

[0023] generating an input vector according to the scene data and the state data;

[0024] Performing prediction processing based on the input vector by the prediction model to obtain a usage probability of each application service in the target carrier;

[0025] Based on the multiple usage probabilities and preset recommendation templates, the recommendation scheme including one or more recommended applications is determined; wherein the recommendation template includes: a probability threshold for screening the application services, the recommendation type of the recommended application, the number of the recommended applications under each recommendation type, and the order of the recommended applications.

[0026] In the above implementation process, when performing recommendation processing, the current scene data and state data of the target carrier can be obtained, and an input vector can be generated based on the scene data and state data. The input vector is used as the input data of the prediction model for calculation and processing to obtain the usage probability of each application service in the target carrier. In order to provide the target carrier with a variety of different types of recommended applications, corresponding recommendation templates can be preset to screen multiple application services with high usage probabilities and corresponding recommendation types as recommended applications based on the usage probabilities and recommendation templates, and the recommended applications can be sorted and processed to obtain corresponding recommendation solutions. It is possible to predict the usage probability of application services based on the current actual situation, and to screen and process application services in combination with the corresponding recommendation templates to obtain recommendation solutions that are relevant to the usage needs of the target carrier, thereby improving the effectiveness of the recommendation solutions and achieving targeted personalized recommendations.

[0027] Optionally, the recommended template is determined by:

[0028] Determining selection preference information of the target carrier based on historical application data of the target carrier; wherein the selection preference information includes: selected service type, application quantity and application distribution;

[0029] Based on the selection preference information, the recommendation template is generated.

[0030] In the above implementation process, considering that users' needs for recommended applications vary across different target carriers, the target carrier's historical application data can be used to determine user preference information regarding application selection, such as service type, number of applications, and application distribution, to generate a corresponding recommendation template based on this preference information. Generating a corresponding recommendation template based on the user's historical application selections within the target carrier effectively improves the correlation between the recommendation template and the user's selection habits, thereby increasing the rationality of the arrangement of recommended applications within the recommendation template and thereby increasing the probability that the user will select and use the recommended application.

[0031] Optionally, the method further includes:

[0032] monitoring the scene data and the status data of the target carrier;

[0033] If it is determined that the scene data and / or the state data have changed, acquiring the change data of the target carrier;

[0034] The recommendation solution is updated according to the change data to obtain an updated recommendation solution.

[0035] In the above implementation process, since the scene conditions and status conditions corresponding to the target vehicle are subject to change, the scene data and status data of the target vehicle can be monitored in real time. When changes in the scene data and / or status data are detected, the target vehicle's change data can be obtained in real time, and the generated recommendation plan can be updated based on the change data, and the updated recommendation plan can be re-recommended. The ability to dynamically update the recommendation plan in real time based on the actual changes in the target vehicle effectively improves the real-time performance, flexibility, and adaptability of the recommendation plan, meeting the needs of use in various complex driving scenarios.

[0036] Optionally, the method further includes:

[0037] Performing recommendation processing on the recommended solution; wherein the recommendation processing includes recommending voice information and / or recommending display information;

[0038] The target application is started according to the selection information fed back by the target carrier based on the recommendation solution.

[0039] In the above implementation process, after generating a recommended solution, the recommended solution can be recommended through voice, display, or other means, and selection information based on the user's feedback on the recommended solution can be obtained on the target carrier. If the selection information includes a recommended application in the recommended solution, the recommended application is used as the target application and the target application is launched to provide the corresponding service function to the user. The ability to promptly launch the corresponding application to provide the user with the service based on the user's feedback on the recommended solution effectively improves the response speed of the service function and further optimizes the user experience.

[0040] In a second aspect, an embodiment of the present application further provides an electronic device, comprising a memory and a processor, wherein the memory stores program instructions, and when the processor runs the program instructions, it executes the steps of any one of the application recommendation methods.

[0041] In a third aspect, an embodiment of the present application further provides a computer program product, which includes a computer program / instructions, and when the computer program / instructions are executed by a processor, the steps of any one of the application recommendation methods are implemented.

[0042] In summary, the embodiments of the present application provide an application recommendation method, electronic device, and computer program product that can use a prediction model to combine current scene data and status data to make predictions for targeted recommendations, effectively improving the correlation between the recommendation scheme and the user's actual needs, thereby optimizing the user's usage experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0044] Figure 1 A block diagram of an electronic device provided in an embodiment of the present application;

[0045] Figure 2 A flowchart of the first application recommendation method provided in an embodiment of the present application;

[0046] Figure 3 A flowchart of the second application recommendation method provided in an embodiment of the present application;

[0047] Figure 4 A flowchart of the third application recommendation method provided in an embodiment of the present application;

[0048] Figure 5 A flowchart of the fourth application recommendation method provided in an embodiment of the present application;

[0049] Figure 6 A detailed flowchart of step S300 provided in an embodiment of the present application;

[0050] Figure 7 A flowchart of the fifth application recommendation method provided in an embodiment of the present application;

[0051] Figure 8 A flowchart of the sixth application recommendation method provided in an embodiment of the present application;

[0052] Figure 9 A flowchart of the seventh application recommendation method provided in an embodiment of the present application;

[0053] Figure 10 A schematic diagram of the structure of an application recommendation device provided in an embodiment of the present application.

[0054] Icons: 100 - electronic device; 111 - memory; 112 - storage controller; 113 - processor; 114 - peripheral interface; 115 - input and output unit; 116 - display unit; 800 - application recommendation device; 810 - acquisition module; 820 - recommendation module. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the embodiments described are only part of the embodiments of the present application, not all of them. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the embodiments of the present application.

[0056] Current application recommendation functions usually require developers or operators to manually configure the recommendation weights for each type of service application based on their own experience before they are launched. Some existing technologies only determine the recommendation list based on a single factor such as the vehicle's geographical location or weather conditions. For example, navigation applications are recommended when the vehicle enters a highway. As a result, the recommended application solutions are relatively fixed and inflexible, and personalized recommendations cannot be made based on the user's actual needs. As a result, the recommended application services are disconnected from the user's actual needs, and the probability of users using the application services is low, resulting in a poor user experience and an inability to provide effective services to users.

[0057] In order to solve the above problems, an embodiment of the present application provides an application recommendation method, which is applied to electronic devices. The electronic devices can be servers, personal computers (PCs), tablets, smart phones, personal digital assistants (PDAs), and other electronic devices with logical computing functions. It can use a prediction model to combine current scene data and status data to make predictions for targeted recommendations, effectively improving the correlation between the recommendation plan and the user's actual needs, thereby optimizing the user's usage experience.

[0058] Optionally, see Figure 1 , Figure 1 This is a block diagram of an electronic device provided in an embodiment of the present application. The electronic device 100 may include a memory 111, a storage controller 112, a processor 113, a peripheral interface 114, an input and output unit 115, and a display unit 116. It will be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the electronic device 100. For example, the electronic device 100 may further include Figure 1 More or fewer components than shown, or with Figure 1 Different configurations shown.

[0059] The aforementioned memory 111, storage controller 112, processor 113, peripheral interface 114, input / output unit 115, and display unit 116 are electrically connected to each other, directly or indirectly, to enable data transmission or interaction. For example, these components may be electrically connected to each other via one or more communication buses or signal lines. The aforementioned processor 113 is used to execute the executable modules stored in the memory.

[0060] The memory 111 may be, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), etc. The memory 111 is used to store programs, and the processor 113 executes the programs after receiving an execution instruction. The method executed by the electronic device 100 defined by the process disclosed in any embodiment of the present application may be applied to the processor 113 or implemented by the processor 113.

[0061] The processor 113 may be an integrated circuit chip with signal processing capabilities. The processor 113 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The methods, steps, and logic block diagrams disclosed in the embodiments of the present application may be implemented or executed. The general-purpose processor may be a microprocessor or any conventional processor.

[0062] The peripheral interface 114 couples various input / output devices to the processor 113 and the memory 111. In some embodiments, the peripheral interface 114, the processor 113, and the memory controller 112 can be implemented in a single chip. In other embodiments, they can be implemented in separate chips.

[0063] The input / output unit 115 is used to provide input data to the user. The input / output unit 115 can be, but is not limited to, a mouse and a keyboard.

[0064] The above-mentioned display unit 116 provides an interactive interface (such as a user operation interface) between the electronic device 100 and the user or is used to display image data for the user's reference. In this embodiment, the display unit can be a liquid crystal display or a touch display. If it is a touch display, it can be a capacitive touch screen or a resistive touch screen that supports single-point and multi-touch operations. Supporting single-point and multi-touch operations means that the touch display can sense touch operations generated simultaneously from one or more positions on the touch display, and hand over the sensed touch operations to the processor for calculation and processing. In an embodiment of the present application, the display unit 116 can display data such as generated recommendation plans.

[0065] The electronic device in this embodiment can be used to execute each step in each application recommendation method provided in the embodiment of this application. The implementation process of the application recommendation method is described in detail below through several embodiments.

[0066] See also Figure 2 , Figure 2 This is a flowchart of a first application recommendation method provided in an embodiment of the present application. The method may include steps S200-S300.

[0067] Step S200: Acquire current scene data and status data of the target carrier.

[0068] Among them, the target carrier can be various types of vehicles, ships, aircraft and other equipment with mobile driving functions. When making application recommendations, the current scene data and status data of the target carrier can be obtained first. The scene data can include weather data, lane data and distance information between the target carrier and the service point in the current environment of the target carrier, and other data related to the scene environment. The status data can include the current carrier speed, carrier position, carrier temperature, carrier personnel data, hardware status and other data related to the current status of the target carrier.

[0069] It should be noted that scene data and status data can be collected through various types of sensors set on the target carrier, and corresponding scene data can also be obtained through multiple service terminals connected to the target carrier in communication. The scene data and status data are associated based on the time node of collection.

[0070] For example, weather data can include various weather types, such as sunny, light rain, and heavy snow, as well as data such as temperature and humidity. Weather data can be obtained via a weather sensor installed in the target carrier, or via a server connected to the target carrier. For example, the data type of the weather type can include coded information, such as 0 (sunny), 1 (cloudy), 2 (overcast), 3 (light rain), 4 (heavy rain), 5 (snow), 6 (dust), and 7 (smog). Temperature data at the current location can be obtained via a temperature sensor installed externally to the target carrier, or via a server connected to the target carrier. The data type of the temperature data can be a floating point number, in degrees Celsius, ranging from -50 to 60°C, with an example value of 22.5°C. Humidity data at the current time can be obtained via a humidity sensor installed externally to the target carrier. The data type of the humidity data can be a floating point number, ranging from 0 to 100 (representing 0% to 100% humidity), with an example value of 65.3. Lane data may include information related to the driving section, such as the type of road and congestion conditions on which the target carrier is currently traveling. The road type can be determined by matching the carrier position with the corresponding map data. The data type of the road type can be corresponding coded information, for example, 0 (highway), 1 (urban road), 2 (rural road), 3 (parking lot / stationary state), 4 (tunnels, overpasses and other special sections), etc. The road speed limit can be compared with the carrier speed, or the congestion conditions of the current section can be obtained by querying a third-party traffic interface that is connected to the target carrier. The data type of the congestion level can also be coded information, for example, 0 (smooth), 1 (mild congestion), 2 (moderate congestion), 3 (severe congestion), etc. The service point can be a corresponding service area, gas station, charging area, parking lot, restaurant, etc., which can provide the target carrier with a variety of services such as charging, refueling, parking, shopping, dining, shopping, etc. The distance information between the target carrier and the service point can be determined based on the carrier position, driving route, etc. For example, the map service can be called according to the carrier position to extract surrounding facility information, such as the distance to the parking lot, gas station, and restaurant. The distance information can be positioned as the distance value (in km) of the two closest POI (Point of Interest). The data type of the distance information can be set to a floating point number, for example: Parking Distance: floating point number, ranging from 0 to 20 km, Restaurant Distance: floating point number, ranging from 0 to 20 km, etc.

[0071] For example, the corresponding carrier speed can be collected by the speed sensor set on the target carrier. The data type of the carrier speed is a floating point number, and the unit is km / h. For example: range: 0~250km / h, example value: 56.7km / h, acceleration (Acceleration): optional data, derived from the acceleration sensor, the data type can also be a floating point number, and the unit is m / s 2 , for example: Range: -10~10m / s 2 (Negative values ​​indicate deceleration, positive values ​​indicate acceleration), example value: 1.2 m / s 2 The target carrier's location can be acquired through positioning systems such as the GPS (Global Positioning System) and Beidou Positioning System installed on the target carrier. Taking GPS as an example, the carrier's location can include latitude, longitude, and altitude. The latitude and longitude data types are floating point numbers with a range of: -90° to 90° for latitude, with an example of 31.2304°N (N can be omitted for positive values); -180° to 180° for longitude, with an example of 121.4737°E (E can be omitted for positive values). Altitude data is also floating point, in meters, with a range of -500 to +9000 meters, with an example of 30.5 meters. Based on the carrier's location and map data, scenario data such as weather data, lane information, and distance information can be further determined. A temperature sensor can also be installed within the target carrier to collect the internal carrier temperature. The carrier temperature data type can be floating point, with a range of -20° to 50°C, with an example of 24.0°C. Corresponding seat sensors or cameras may also be provided in the target carrier to detect the seating conditions and determine the carrier occupant data in the target carrier. Taking a five-seater car as an example, the data type of the carrier occupant data may be a floating point number ranging from 0 to 5, indicating the number of people currently in the car (including the driver), with an example value of 2 (driver + 1 passenger), etc. The hardware status may include various hardware-related status information such as the target carrier's energy status and audio status. For example, the target carrier's remaining fuel level, remaining battery level, and the volume level of the audio equipment installed in the target carrier may be represented by a floating point number with a data type ranging from 0 to 100 (0 for mute and 100 for maximum volume), with an example value of 12, etc.

[0072] Step S300: Substitute the scene data and the state data into the prediction model corresponding to the target carrier to perform prediction, and obtain a recommendation solution including at least one recommended application.

[0073] Among them, it is possible to use the historical scene data, historical status data and historical application data of the target carrier to build a prediction model for predicting the usage of application services, and use the current data as input data in the input prediction model to make predictions and obtain a recommendation plan containing one or more recommended applications.

[0074] Optionally, multiple recommendation types can be set in the target carrier, such as navigation services, music services, payment services, adjustment services, etc., and multiple different types of application services can also be set under each recommendation type. For example, navigation services include navigation apps from multiple different suppliers, and adjustment services include temperature adjustment applications, seat adjustment applications, etc.

[0075] It should be noted that the prediction model is constructed and trained based on the historical scene data, historical status data and historical application data of the target carrier. The prediction model can be set to a variety of models with prediction functions, such as deep neural network models.

[0076] exist Figure 2 In the embodiment shown, targeted application recommendations can be made based on the current actual situation and historical habits of the target carrier, effectively improving the correlation between the recommendation scheme and the user's actual needs, so as to increase the probability of the target carrier using the recommended application, thereby optimizing the user's experience of using the application service.

[0077] Optionally, see Figure 3 , Figure 3 This is a flow chart of a second application recommendation method provided in an embodiment of the present application. The method may further include steps S410-S430.

[0078] Step S410 , combining historical time nodes, constructing training data and label data based on the acquired historical scene data, historical state data, and historical application data of the target carrier.

[0079] Among them, the associated historical time nodes and various historical data can be combined to construct corresponding training data and label data.

[0080] It should be noted that historical scene data, historical status data and historical application data are associated through historical time nodes.

[0081] Optionally, historical scenario data may include various data related to historical scenarios, such as historical weather data, historical lane data, and historical distance information between the target vehicle and the service point. Historical status data may include data related to the status of the historical vehicle, such as historical vehicle speed, historical vehicle location, historical vehicle temperature, historical vehicle personnel data, and historical hardware status. Historical application data may include data related to the usage of historical applications, such as historical application identity information and historical application usage information of historical applications selected by users. The collection method, data structure, and data type of historical scenario data and historical status data are similar to those of scenario data and are not further described here.

[0082] For example, historical application data can be extracted from historical application usage records, which can be stored in the target carrier's control system's historical behavior log to record the user's selection or rejection of each recommended application. Historical application identity information can include unique identification information such as the ID and name of the used application, and can include a string of a certain length, for example, App1 = 1001, App2 = 1002, App3 = 1003, etc. Historical application usage information can include various data such as the usage duration, usage frequency, and scenario-related behavior of each historical application. The usage duration data type can be a floating point number in minutes, ranging from 0 to 300 (until a single use is interrupted). For example, over a period of time, such as 7 days, the total usage duration of App1 is 120 minutes, App2 is 200 minutes, and App3 is 45 minutes. The usage frequency data type can be a count or frequency value, for example, the number of uses in a week ranges from 0 to 100, for example: App1: 20 times / week, App2: 15 times / week, App3: 5 times / week. Scenario-related behavior can be used to compile statistics on users' application tendencies in specific scenarios. For example, under the conditions of "urban roads + moderate congestion + light rain", users used to often use music application services (with high frequency and long usage duration), while under the conditions of "highways + sunny days", they used navigation application services more often.

[0083] Optionally, multiple historical scenario data, historical status data, and historical application data can be integrated to obtain corresponding feature data. For example, scenario Key: (road type = 1, congestion condition = 2, weather type = 3), the corresponding user's frequently used App1 probability is 0.8, App2 probability is 0.15, App3 probability is 0.05, etc. Integrate multiple feature data into a high-dimensional vector input model to obtain the corresponding feature vector X as training data. For example, X = [Speed, Acceleration, Latitude, Longitude, Road Type, Traffic Level, Weather, Temperature, Cabin Temperature, Volume Level, Seat Occupancy, Parking Distance, Restaurant Distance, Time Of Day, Day Of Week, App Usage Stats], that is, X = [historical carrier speed, historical carrier acceleration, historical carrier latitude, historical carrier longitude, historical road type, historical traffic congestion, historical weather type, historical external temperature, historical internal temperature, historical volume level, historical carrier personnel data, historical parking distance information, historical restaurant distance information, historical time nodes (time period, number of days), historical application data], where Time Of Day in the historical time node is the time period within a day, and an integer from 0 to 23 can be used to represent the number of hours. Example value: 15 (representing 15:00 to 15:59), Day Of Week can be 1 to 7 to represent Monday to Sunday, example value is 3 (Wednesday), etc.

[0084] Optionally, for features related to user application preferences, the frequency and duration of use of the top N applications used by the user in the past period of time, such as 7 days, can be used as features, or scenario-based preference values ​​can be used as part of the feature vector. For example, the following characteristics can be added: App1 usage frequency = 20 times / week, App2 usage frequency = 15 times / week, App3 usage frequency = 5 times / week, or scenario preference distribution can be added, such as APP1 usage probability = 0.8, APP2 usage probability = 0.05, APP3 usage probability = 0.15, etc. In this way, the final feature vector may have dozens to hundreds of dimensions, each dimension has a clear value range and meaning, so as to improve the effectiveness of the training data.

[0085] It should be noted that the label data can be determined based on historical scenario data, historical status data and historical application data. For example, in the case of a certain feature vector X, if APP1 is used, the corresponding label is 1, and if APP2 is not used, the corresponding label is 0, so that use and non-use are marked and distinguished by 1 and 0.

[0086] Optionally, the training data can be within the past six months to improve the validity of the training data. The training data can be split into a training set, a validation set, and a test set in a ratio of 8:1:1. The initial prediction model is iteratively trained using an optimizer (e.g., Adam with a learning rate lr = 0.001), a loss function (e.g., cross entropy), and a number of training rounds (e.g., 10-50 rounds, which can be optimized based on the validation set metrics).

[0087] Step S420: training the initial prediction model based on the training data, and verifying the prediction results of the initial prediction model based on the label data.

[0088] The training data can be used to train the initial prediction model, and the label data can be used to verify the prediction results of the initial prediction model during the training process.

[0089] It should be noted that the initial prediction model can be any of a variety of models capable of predicting usage probability, such as a corresponding deep neural network model. The input to the initial prediction model is the multiple feature vectors X from the training data, and the output is the usage probability or score for each application service provided in the target carrier. For example, if the target carrier has M recommended application services, the model outputs a prediction result of length M, a vector Y, where each element Yi represents the predicted probability of use for that application. For each application Appi, the output Yii is in the range [0, 1], for example: YApp1 = 0.7, Yapp2 = 0.2, Yapp3 = 0.1, etc.

[0090] Optionally, the accuracy of the prediction result can be verified based on the set label data. For example, if the label corresponding to the application service with a higher probability of being used is 1, it indicates that the prediction is accurate. If the label corresponding to the application service with a higher probability of being used is 0, it indicates that the prediction is wrong, so as to calculate the corresponding prediction accuracy, which can be percentage data.

[0091] Step S430: When the prediction accuracy of the prediction result is greater than or equal to a preset accuracy threshold, a prediction model is obtained.

[0092] Among them, when the prediction accuracy of the prediction result is greater than or equal to the preset accuracy threshold, it indicates that the current prediction accuracy of the initial prediction model is high and meets the usage requirements. The training can be stopped and the current initial prediction model can be used as the prediction model.

[0093] For example, the accuracy threshold can be set to a corresponding percentage threshold, such as 50%. When the prediction accuracy is greater than or equal to 50%, it indicates that the current prediction accuracy of the initial prediction model is high and meets the prediction requirements during use.

[0094] exist Figure 3 In the embodiment shown, the corresponding prediction model can be trained based on historical scenario conditions, historical status conditions and historical usage of the application, and the prediction accuracy of the prediction model can be effectively improved by verifying the prediction results, further optimizing the correlation between the recommended solutions determined using the prediction model and the actual demand conditions.

[0095] Optionally, see Figure 4 , Figure 4 This is a flow chart of a third application recommendation method provided in an embodiment of the present application. The method may further include steps S441-S442.

[0096] Step S441 : determining a plurality of different training weights according to different time periods corresponding to the training data.

[0097] It should be noted that due to changes in time, seasons, etc. will also change. Users' usage habits in different seasons, such as temperature habits, speed habits, etc., therefore, considering that the usage habits of users in the target carrier will change accordingly over time, in the use of the prediction model, multiple training weights of different sizes can be determined based on the training data of different time periods. It should be noted that the training weight is inversely proportional to the time distance between the time period and the current time node, that is, the training weight corresponding to the time period close to the current time node is larger, and the training weight of the time period far from the current time node is smaller. For example, the training weight of training data within 1 month from the current time node is 0.5, the training weight of training data within one month and six months from the current time node is 0.3, and the training weight of training data within six months and twelve months from the current time node is 0.2.

[0098] Step S442: updating the prediction model based on the preset update cycle and training weight.

[0099] Among them, the corresponding update cycle can be set in advance, such as one week or one month, and the update cycle can be set and adjusted according to the actual needs of the user. The prediction model can be iteratively updated or retrained in combination with the set update cycle and training weight.

[0100] For example, the prediction model can be updated in the target carrier, or in the server to which the target carrier is communicatively connected, with the target carrier downloading the updated model via the network. If the target carrier or server has high computing performance, the initial prediction model can be retrained at the beginning of each update cycle based on training data with different training weights to obtain an updated prediction model, thereby further improving the effectiveness of the prediction model. If the computing performance of the target carrier or server is average, the original prediction model can be iteratively trained based on training data with different training weights to optimize the parameters therein and obtain an updated prediction model.

[0101] exist Figure 4 In the embodiment shown, the prediction model can be updated based on recent training data, thereby effectively sensing changes in user usage habits and periodically updating the prediction model in real time based on changes in habits, thereby improving the real-time performance and effectiveness of the prediction model.

[0102] Optionally, see Figure 5 , Figure 5 This is a flow chart of the fourth application recommendation method provided in an embodiment of the present application. The method may further include steps S451-S452.

[0103] Step S451: Acquire recent historical data within a preset update period before the current time node.

[0104] Among them, during the use of the prediction model, a corresponding update cycle can be set for regular updates, such as one week or one month. When updating, recent historical data within the update cycle before the current time node can be obtained, for example, data within the previous week can be obtained as recent historical data.

[0105] Step S452: Update the prediction model based on recent historical data and update cycle.

[0106] Among them, recent historical data can be used as training data to iteratively update the current prediction model to optimize various parameters within the prediction model. The computing power cost required for the update is relatively small, and the time required for the update is relatively short.

[0107] For example, the prediction model may be updated in the target carrier, or in a server to which the target carrier is communicatively connected, with the target carrier downloading the updated model through a network.

[0108] exist Figure 5 In the illustrated embodiment, the prediction model can be periodically updated using recent historical data within each update period as training data, effectively improving the updating efficiency and effectiveness of the prediction model.

[0109] Optionally, see Figure 6 , Figure 6 A detailed flow chart of step S300 is provided in an embodiment of the present application. Step S300 may include steps S310-S330.

[0110] Step S310: Generate an input vector according to the scene data and the state data.

[0111] Among them, when performing recommendation processing, the current scene data and state data of the target carrier can be obtained, and an input vector can be generated based on the scene data and state data. The input vector can be recorded as X1. The data structure of the input vector is similar to that of the feature vector and will not be repeated here.

[0112] Step S320 , performing prediction processing based on the input vector by using a prediction model to obtain the usage probability of each application service in the target carrier.

[0113] The input vector can be used as input data of the prediction model for calculation and processing to obtain the probability of each application service being used. For example, when there are i application services, the prediction result of the prediction model is Y1-Yi.

[0114] Step S330 : determining a recommendation solution including one or more recommended applications based on the multiple usage probabilities and the preset recommendation templates.

[0115] Among them, taking into account the diversity of recommendation methods, corresponding recommendation templates can be set in advance. The recommendation templates may include: probability thresholds for screening application services, recommendation types of recommended applications, the number of recommended applications under each recommendation type, and the order of recommended applications, so as to screen multiple application services with high usage probabilities and corresponding recommendation types as recommended applications based on the usage probability and recommendation templates, and sort the recommended applications to obtain corresponding recommendation plans.

[0116] For example, the recommendation type is the service type of the application service, such as navigation, payment, music, etc. The recommendation scheme may include a corresponding list, for example, selecting the top three application services with the highest probability of being used as recommended applications, and marking the recommendation type of each recommended application, or selecting two application services with the highest probability of being used in each recommendation type as recommended applications, etc. When recommending, multiple recommended applications are arranged from high to low according to the probability of being used.

[0117] For example, when generating recommendations, the recommended service type can be determined based on the target vehicle's scenario. For example, when driving on a highway, navigation, music, or communication apps may be prioritized, while near a parking lot, the system may recommend parking reservations or payment apps. Based on the recommended types, the recommendation list can be further refined based on user preferences. For example, when recommending music apps, a specific music player or radio station may be recommended based on the user's historical habits.

[0118] Optionally, each recommended application in the recommendation scheme may include data such as the corresponding application name.

[0119] exist Figure 6 In the embodiment shown, the probability of application service usage can be predicted based on the current actual situation, and the application service can be screened and processed in combination with the corresponding recommendation template to obtain a recommendation plan that is relevant to the usage needs of the target carrier, thereby improving the effectiveness of the recommendation plan and achieving targeted personalized recommendations.

[0120] Optionally, see Figure 7 , Figure 7 This is a flowchart of the fifth application recommendation method provided in an embodiment of the present application. The method may further include steps S510-S520.

[0121] Step S510: determining selection preference information of the target carrier according to historical application data of the target carrier.

[0122] Among them, considering that users have different needs for recommended applications in different target carriers, the selection preference information such as service type, number of applications and application distribution of users in the target carrier can be determined based on the historical application data of the target carrier.

[0123] Optionally, the selection preference information may include: the selected service type, the number of applications, and the distribution of applications. The selected service type may be whether the user is accustomed to selecting from multiple application services within the same service type, or from multiple application services within multiple different service types. The number of applications may be the recommended number that the user is accustomed to, for example, the user is accustomed to selecting from three application services, or the user is accustomed to selecting the one application service with the highest probability of being used, etc. The application distribution may include the user's accustomed layout mode, for example, if the user is accustomed to processing multiple application services of different service types in a unified manner, the multiple application services are arranged together, or if the user is accustomed to processing multiple application services of different service types separately, the multiple application services are arranged separately, etc.

[0124] Step S520: Generate a recommendation template based on the selection preference information.

[0125] Among them, multiple selection preference information can be used in combination with deep learning models for training processing to obtain corresponding recommendation templates.

[0126] exist Figure 7 In the embodiment shown, a corresponding recommendation template can be generated based on the user's historical application selection in the target carrier, effectively improving the correlation between the recommendation template and the user's selection habits, so as to improve the rationality of the arrangement of recommended applications in the recommendation template, thereby increasing the probability of users selecting recommended applications for use.

[0127] Optionally, see Figure 8 , Figure 8 This is a flowchart of the sixth application recommendation method provided in an embodiment of the present application. The method may further include steps S610-S630.

[0128] Step S610: monitoring the scene data and status data of the target carrier.

[0129] Since the scene conditions and status conditions corresponding to the target carrier are changing, the scene data and status data of the target carrier may be monitored in real time to monitor the changes.

[0130] For example, monitoring can be performed by periodic data collection and comparison, for example, data collection is performed every 2 seconds, and the latest data is compared with the previous data. If the data changes, it indicates that the condition of the target carrier has changed.

[0131] Step S620: If it is determined that the scene data and / or state data has changed, the change data of the target carrier is obtained.

[0132] When a change in scene data and / or state data is detected, the change data of the target carrier can be acquired in real time. The change data may be the scene data and / or state data after the change in the target carrier.

[0133] Step S630: Update the recommended solution according to the changed data to obtain an updated recommended solution.

[0134] Among them, the changed data can be processed to obtain the corresponding input vector and input into the prediction model to obtain an updated recommendation plan, or the original recommendation plan can be directly updated according to the difference between the changed data and the unchanged data to obtain an updated recommendation plan for re-recommendation processing.

[0135] For example, if the target vehicle changes as follows: it has just entered the vicinity of a parking lot and the Parking Distance is less than 0.3km, then navigation applications can be added to the recommended solution, or the priority of navigation applications can be advanced. If the target vehicle changes as follows: the weather suddenly changes from sunny to rainy, then weather applications or traffic query applications can be added to the recommended solution, or the priority of weather applications or traffic query applications can be advanced. If the target vehicle changes as follows: the driving status of the target vehicle changes from long-distance driving to low-speed driving in the city, then the previous navigation applications can be replaced with parking applications in the recommended solution.

[0136] Optionally, when the user uses APP1 for a long time or frequently, the priority of APP1 may be lowered in subsequent recommendation schemes to encourage diversified recommendations.

[0137] exist Figure 8 In the illustrated embodiment, the recommended solution can be dynamically updated in real time according to the actual changes in the target carrier, effectively improving the real-time performance, flexibility, and adaptability of the recommended solution and meeting the usage requirements in various complex driving scenarios.

[0138] Optionally, see Figure 9 , Figure 9 This is a flowchart of the seventh application recommendation method provided in an embodiment of the present application. The method may further include steps S710-S720.

[0139] Step S710: Perform recommendation processing on the recommended solution.

[0140] Among them, the recommendation processing may include recommendation voice information and / or recommendation display information. After the recommendation plan is generated, the recommendation plan can be recommended through voice, display, etc. For example, the human-computer interaction interface of the recommendation plan can be displayed through a display device such as a display screen set in the target carrier, or the recommendation plan can be played through the audio set in the target carrier.

[0141] Step S720 : starting the target application according to the selection information fed back by the target carrier based on the recommendation solution.

[0142] Among them, the selection information of the user on the target carrier based on the recommendation scheme feedback is obtained. When the selection information includes the recommended application in the recommendation scheme, the recommended application is used as the target application, and the target application is started to provide the user with corresponding service functions.

[0143] For example, when the recommendation is processed as recommendation voice information, the selection information can be obtained by recognizing the user's voice. When the recommendation is processed as recommendation display information, the selection information can be obtained through buttons and keys clicked by the user on the interactive interface. The selection information can also be obtained through voice recognition. For example, the target carrier prompts: "It is detected that you are about to arrive at your destination. Do you need recommended parking services?" The user can choose whether to accept the recommended application through voice commands.

[0144] It should be noted that users can select recommended applications from the recommendation scheme, or select other non-recommended application services according to their own needs. The selection information can be parsed to determine the user's actual selection target, thereby starting the target application required by the user.

[0145] exist Figure 9 In the illustrated embodiment, corresponding applications can be launched in a timely manner to provide services to users based on user feedback on recommended solutions, thereby effectively improving the response speed of service functions and further optimizing the user experience.

[0146] See also Figure 10 , Figure 10 This is a schematic diagram of the structure of an application recommendation device provided in an embodiment of the present application. The application recommendation device 800 may include a collection module 810 and a recommendation module 820;

[0147] The acquisition module 810 is used to obtain the current scene data and status data of the target carrier; wherein the scene data includes at least one of current weather data, lane data, and distance information between the target carrier and the service point; the status data includes at least one of current carrier speed, carrier position, carrier temperature, carrier personnel data, and hardware status;

[0148] The recommendation module 820 is used to substitute the scene data and status data into the prediction model corresponding to the target carrier for prediction, and obtain a recommendation plan including at least one recommended application; wherein the prediction model is constructed based on the historical scene data, historical status data and historical application data of the target carrier.

[0149] In an optional embodiment, the application recommendation device 800 may further include a modeling module for constructing training data and label data based on the historical scene data, historical status data and historical application data of the target carrier obtained in combination with historical time nodes; wherein the historical scene data, historical status data and historical application data are associated through historical time nodes; the historical scene data include historical weather data, historical lane data and at least one of historical distance information between the target carrier and the service point; the historical status data include at least one of historical carrier speed, historical carrier position, historical carrier temperature, historical carrier personnel data and historical hardware status; the historical application data include historical application identity information and historical application usage information of historical applications selected by users; the initial prediction model is trained based on the training data, and the prediction result of the initial prediction model is verified based on the label data; and the prediction model is obtained when the prediction accuracy of the prediction result is greater than or equal to the preset accuracy threshold.

[0150] In an optional embodiment, the application recommendation device 800 may further include a first model updating module for determining a plurality of different training weights based on different time periods corresponding to the training data; wherein the training weight is inversely proportional to the time distance between the time period and the current time node; and the prediction model is updated based on a preset update cycle and training weight.

[0151] In an optional embodiment, the application recommendation device 800 may further include a second model updating module for obtaining recent historical data within a preset update period before the current time node; and updating the prediction model based on the recent historical data and the update period.

[0152] In an optional embodiment, the recommendation module 820 is specifically used to: generate an input vector based on scenario data and status data; perform prediction processing based on the input vector through a prediction model to obtain the probability of use of each application service in the target carrier; based on multiple usage probabilities and a preset recommendation template, determine a recommendation scheme containing one or more recommended applications; wherein the recommendation template includes: a probability threshold for screening application services, a recommendation type of recommended applications, the number of recommended applications under each recommendation type, and the order of recommended applications.

[0153] In an optional embodiment, the application recommendation device 800 may further include a template module for determining selection preference information of the target carrier based on the historical application data of the target carrier; wherein the selection preference information includes: the selected service type, the number of applications and the distribution of applications; and generating a recommendation template based on the selection preference information.

[0154] In an optional embodiment, the recommendation module 820 is also used to: monitor the scene data and status data of the target carrier; if it is determined that the scene data and / or status data have changed, obtain the change data of the target carrier; update the recommendation plan according to the change data to obtain an updated recommendation plan.

[0155] In an optional embodiment, the recommendation module 820 is further used to: perform recommendation processing on the recommendation scheme; wherein the recommendation processing includes recommending voice information and / or recommending display information; and start the target application according to the selection information fed back by the target carrier based on the recommendation scheme.

[0156] Since the principle of solving the problem by the application recommendation device 800 in the embodiment of the present application is similar to that in the embodiment of the aforementioned application recommendation method, the implementation of the application recommendation device 800 in this embodiment can refer to the description in the embodiment of the aforementioned application recommendation method, and the repeated parts will not be repeated.

[0157] An embodiment of the present application also provides a computer program product, which includes a computer program / instructions. When the computer program / instructions are executed by a processor, the steps of any one of the application recommendation methods provided in this embodiment are implemented.

[0158] In the several embodiments provided in this application, it should be understood that the disclosed devices can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices according to the multiple embodiments of the present application. In this regard, each box in the block diagram can represent a module, a program segment or a part of a code, and the module, program segment or a part of the code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram, and the combination of the block diagrams, can be implemented with a dedicated hardware-based system that performs the specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0159] In addition, the functional modules in each embodiment of the present application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0160] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0161] The foregoing is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. Various modifications and variations are possible for those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included within the scope of protection of the present application. It should be noted that similar reference numerals and letters represent similar items in the following figures. Therefore, once an item is defined in one figure, it does not need to be further defined or explained in subsequent figures.

[0162] The above is only a specific implementation method of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed in this application, which should be covered by the scope of protection of the present application.

[0163] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, the elements defined by the phrase "comprising..." do not exclude the presence of other identical elements in the process, method, article, or device comprising the elements.

Claims

1. An application recommendation method, characterized in that: The method comprises: Acquiring current scene data and status data of the target carrier; wherein the scene data includes at least one of current weather data, lane data, and distance information between the target carrier and the service point; and the status data includes at least one of current carrier speed, carrier position, carrier temperature, carrier personnel data, and hardware status; Substituting the scenario data and the status data into a prediction model corresponding to the target carrier for prediction, a recommendation solution including at least one recommended application is obtained; wherein the prediction model is constructed based on historical scenario data, historical status data and historical application data of the target carrier.

2. The method according to claim 1, characterized in that in, The prediction model is constructed in the following way: In conjunction with historical time nodes, training data and label data are constructed based on the historical scene data, historical status data, and historical application data of the target carrier obtained; wherein the historical scene data, the historical status data, and the historical application data are associated through the historical time nodes; the historical scene data includes at least one of historical weather data, historical lane data, and historical distance information between the target carrier and the service point; the historical status data includes at least one of historical carrier speed, historical carrier position, historical carrier temperature, historical carrier personnel data, and historical hardware status; and the historical application data includes historical application identity information and historical application usage information of historical applications selected by users; Training an initial prediction model based on the training data, and verifying a prediction result of the initial prediction model based on the label data; When the prediction accuracy of the prediction result is greater than or equal to a preset accuracy threshold, the prediction model is obtained.

3. The method according to claim 2, characterized in that The method further comprises: Determining a plurality of different training weights according to different time periods corresponding to the training data; wherein the training weight is inversely proportional to the time distance between the time period and the current time node; The prediction model is updated based on a preset update cycle and the training weights.

4. The method according to claim 2, characterized in that The method further comprises: Get the recent historical data within the preset update period before the current time node; The prediction model is updated based on the recent historical data and the update cycle.

5. The method according to any one of claims 1 to 4, characterized in that Substituting the scenario data and the state data into a prediction model corresponding to the target carrier to perform prediction, and obtaining a recommendation solution including at least one recommended application, includes: generating an input vector according to the scene data and the state data; Performing prediction processing based on the input vector by the prediction model to obtain a usage probability of each application service in the target carrier; Based on the multiple usage probabilities and preset recommendation templates, the recommendation scheme including one or more recommended applications is determined; wherein the recommendation template includes: a probability threshold for screening the application services, the recommendation type of the recommended application, the number of the recommended applications under each recommendation type, and the order of the recommended applications.

6. The method according to claim 5, characterized in that in, The recommended template is determined in the following way: Determining selection preference information of the target carrier based on historical application data of the target carrier; wherein the selection preference information includes: selected service type, application quantity and application distribution; Based on the selection preference information, the recommendation template is generated.

7. The method according to any one of claims 1 to 4, characterized in that The method further comprises: monitoring the scene data and the status data of the target carrier; If it is determined that the scene data and / or the state data have changed, acquiring the change data of the target carrier; The recommendation solution is updated according to the change data to obtain an updated recommendation solution.

8. The method according to any one of claims 1 to 4, characterized in that The method further comprises: Performing recommendation processing on the recommended solution; wherein the recommendation processing includes recommending voice information and / or recommending display information; The target application is started according to the selection information fed back by the target carrier based on the recommendation solution.

9. An electronic device, characterized in that: The electronic device includes a memory and a processor, wherein program instructions are stored in the memory, and when the processor runs the program instructions, the steps in the method according to any one of claims 1 to 8 are executed.

10. A computer program product, characterized in that The computer program product comprises a computer program / instruction, and when the computer program / instruction is executed by a processor, the steps in the method according to any one of claims 1 to 8 are implemented.

Images