Model training method, vehicle service determination method, device, medium and equipment

Abstract

The present disclosure relates to a model training method, a vehicle service determination method, an apparatus, a medium and an equipment. The method comprises: obtaining vehicle data corresponding to a vehicle and historical operation data provided by a user; performing feature dimension extraction according to the vehicle data and the historical operation data to obtain a target dimension corresponding to an input feature of a service training sample; performing feature extraction from the vehicle data and the historical operation data according to the target dimension to obtain a service input feature and a service result in the service training sample, wherein the service result comprises starting a target service and rejecting the target service; taking the service input feature as an input of a service determination model, taking a service result corresponding to the service input feature as a target output of the service determination model, training the service determination model, and determining a trained service determination model as a service determination model corresponding to the target service.

Description

Technical Field

[0001] This disclosure relates to vehicles, and more specifically, to a model training method, a vehicle service determination method, an apparatus, a medium, and equipment. Background Technology

[0002] As vehicles become increasingly intelligent, they can recommend more and more convenient services to users. Currently, in-vehicle service recommendations mainly rely on expert rules, which involve collecting relevant in-vehicle signal data and matching it based on experience-based rules to recommend in-vehicle services (such as air conditioning temperature control, seat control, window control, and other vehicle control scenarios) to the owner. However, the recommendation model in this process is based on inherent judgments made through experience and rules, making it difficult to achieve personalized recommendations for users and even more difficult to adapt to complex scenarios. Summary of the Invention

[0003] This summary section is provided to briefly introduce the concepts, which will be described in detail in the detailed description section below. This summary section is not intended to identify key or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0004] In a first aspect, this disclosure provides a method for training a vehicle service determination model, the method comprising:

[0005] Obtain vehicle data corresponding to the vehicle and historical operation data provided by the user;

[0006] Based on the vehicle data and the historical operation data, feature dimensions are extracted to obtain the target dimension corresponding to the input features of the service training sample;

[0007] Based on the target dimension, features are extracted from the vehicle data and the historical operation data to obtain service input features and service results in the service training sample, wherein the service results include enabling the target service and rejecting the target service;

[0008] The service input features are used as input to the service determination model, and the service results corresponding to the service input features are used as the target output of the service determination model. The service determination model is trained, and the trained service determination model is determined as the service determination model corresponding to the target service.

[0009] Secondly, this disclosure provides a method for determining vehicle services, the method comprising:

[0010] Obtain the target vehicle data corresponding to the target vehicle and the historical operation data provided by the target user corresponding to the target vehicle;

[0011] Based on the target dimension in the service determination model corresponding to the target service, feature extraction is performed from the target vehicle data and the historical operation data provided by the target user to obtain the current service features of the target vehicle. The service determination model of the target service is obtained based on the training method of the service determination model described in the first aspect.

[0012] Based on the current service characteristics and the service determination model, determine whether the target service is a recommended service output by the target vehicle.

[0013] Thirdly, this disclosure provides a training apparatus for a vehicle service determination model, the apparatus comprising:

[0014] The first acquisition module is used to acquire vehicle data corresponding to the vehicle and historical operation data provided by the user.

[0015] The first processing module is used to extract feature dimensions based on the vehicle data and the historical operation data to obtain the target dimension corresponding to the input features of the service training sample.

[0016] The first extraction module is used to extract features from the vehicle data and the historical operation data according to the target dimension to obtain service input features and service results in the service training sample, wherein the service results include enabling the target service and rejecting the target service;

[0017] The training module is used to train the service determination model by taking the service input features as input, the service results corresponding to the service input features as the target output of the service determination model, and determining the trained service determination model as the service determination model corresponding to the target service.

[0018] Fourthly, this disclosure provides a vehicle service determination device, the device comprising:

[0019] The second acquisition module is used to acquire target vehicle data corresponding to the target vehicle and historical operation data provided by the target user corresponding to the target vehicle.

[0020] The second extraction module is used to extract features from the target vehicle data and the historical operation data provided by the target user based on the target dimension in the service determination model corresponding to the target service, and obtain the current service features of the target vehicle. The service determination model of the target service is obtained based on the training method of the service determination model described in the first aspect.

[0021] The determination module is used to determine whether the target service is a recommended service output by the target vehicle based on the current service characteristics and the service determination model.

[0022] Fifthly, this disclosure provides a computer-readable medium having a computer program stored thereon, which, when executed by a processing device, implements the steps of the method described in the first or second aspect.

[0023] Sixthly, this disclosure provides an electronic device, comprising:

[0024] A storage device on which computer programs are stored;

[0025] A processing device for executing the computer program in the storage device to implement the steps of the method described in the first or second aspect.

[0026] The above technical solution allows users' historical operation records to be used as input to train and construct a service determination model. This enables the trained model to provide personalized services to different users, establishing individual characteristics for each user and improving the personalized recommendation of vehicle services, thus increasing the matching degree between the determined service and the current user. Furthermore, this solution utilizes machine learning methods to model vehicle data, fully leveraging vehicle data and pre-selecting the input feature dimensions of the service determination model. This improves the efficiency and accuracy of model training, thereby enhancing the accuracy of the trained service determination model and enabling effective vehicle service recommendations for users.

[0027] Other features and advantages of this disclosure will be described in detail in the following detailed description section. Attached Figure Description

[0028] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the originals and elements are not necessarily drawn to scale. In the drawings:

[0029] Figure 1 This is a flowchart of a training method for a vehicle service determination model provided according to one embodiment of the present disclosure;

[0030] Figure 2 This is a flowchart illustrating an exemplary implementation of extracting feature dimensions from the vehicle data and the historical operation data to obtain the target dimension corresponding to the input features of the service training sample;

[0031] Figure 3 This is a block diagram of a training apparatus for a vehicle service determination model provided according to one embodiment of the present disclosure;

[0032] Figure 4A schematic diagram of the structure of an electronic device suitable for implementing embodiments of the present disclosure is shown. Detailed Implementation

[0033] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0034] It should be understood that the steps described in the method embodiments of this disclosure may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of this disclosure is not limited in this respect.

[0035] The term "comprising" and its variations as used herein are open-ended inclusions, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below.

[0036] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0037] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0038] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0039] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0040] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the software or hardware, such as the electronic device, application, server, or storage medium performing the operations of this disclosed technical solution, based on the prompt message.

[0041] As an optional but non-limiting implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0042] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this disclosure.

[0043] Meanwhile, it is understood that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and related provisions.

[0044] Figure 1 The diagram shows a flowchart of a training method for a vehicle service determination model according to one embodiment of this disclosure. Figure 1 As shown, the method may include:

[0045] In step 11, vehicle data corresponding to the vehicle and historical operation data provided by the user are obtained.

[0046] Vehicle data can be acquired through various sensors installed on the vehicle. For example, vehicle data may include, but is not limited to, interior and exterior temperature, humidity, driving time, air quality, and the degree to which windows are open or closed. Historical operation data can be information on user operations on relevant vehicle components based on their own habits; this data can be obtained by the user with authorization. Taking the target service as air conditioning temperature control as an example, historical operation data may include, but is not limited to, records of user operations on the air conditioning, windows, seats, etc., to help determine the user's operating habits.

[0047] In step 12, feature dimensions are extracted based on vehicle data and historical operation data to obtain the target dimension corresponding to the input features of the service training sample.

[0048] In step 13, features are extracted from vehicle data and historical operation data according to the target dimension to obtain service input features and service results in the service training samples, wherein the service results include enabling the target service and rejecting the target service.

[0049] The target dimension is a dimension determined to be related to whether the user has enabled the target service. Further extracting relevant data under the target dimension from the acquired data can improve the matching degree between the data in the service training sample obtained based on this and the target service.

[0050] In step 14, the service input features are used as the input to the service determination model, and the service results corresponding to the service input features are used as the target output of the service determination model. The service determination model is trained, and the trained service determination model is determined as the service determination model corresponding to the target service.

[0051] In this step, service input features can be input into the service determination model to obtain the prediction results corresponding to the service determination model. Furthermore, the prediction error of the service determination model can be determined based on the prediction results and service results. Based on the prediction error, the parameters in the service determination model can be adjusted using the gradient descent method to update and train the service determination model.

[0052] For example, training can end when the loss corresponding to the service-determined model is less than the loss threshold, or when the number of training iterations of the service-determined model reaches the threshold. Otherwise, the training process can be repeated to perform multiple training iterations.

[0053] Therefore, the above technical solution allows users' historical operation records to be used as input to train and construct a service determination model. This enables the trained model to provide personalized services to different users, establishing individual characteristics for each user, thereby improving the personalized recommendation of vehicle services and increasing the matching degree between the determined services and the current user. Furthermore, this solution utilizes machine learning methods to model vehicle data, fully leveraging vehicle data and pre-selecting the input feature dimensions of the service determination model. This improves the efficiency and accuracy of model training, ultimately enhancing the accuracy of the trained service determination model and providing effective vehicle service recommendations for users.

[0054] In one possible embodiment, an exemplary implementation of extracting feature dimensions based on the vehicle data and the historical operation data to obtain the target dimension corresponding to the input features of the service training sample is as follows: Figure 2 As shown, this step may include:

[0055] In step 21, the various service input dimensions corresponding to the vehicle data and historical operation data are determined.

[0056] As an example, dimensional analysis can be directly performed on vehicle data and historical operation data to obtain the various dimensions contained in the vehicle data and historical operation data respectively. For example, analyzing vehicle data according to the sensor source corresponding to the vehicle data can determine that it includes temperature, humidity, air quality, and the degree of window opening and closing. In the training phase, vehicle data can also include user feedback operations on services in the vehicle, such as users activating or rejecting a target service. At the same time, the vehicle's driving time can also be recorded, and the obtained dimensions can be used as the service input dimensions corresponding to the vehicle data.

[0057] Historical operation data can be obtained, showing user actions on various objects within a vehicle over a historical period. This record can then be used as the service input dimension for the target service. For example, the historical period can be set according to the actual application scenario, such as operation records from the most recent 7 days.

[0058] For example, by analyzing historical operation data and determining that it includes operation records of objects such as air conditioning, windows, and seats in a vehicle, it can be identified as the service input dimension corresponding to the historical operation data under the target service.

[0059] As another example, an exemplary implementation of determining the various service input dimensions corresponding to the vehicle data and the historical operation data is as follows, which may include:

[0060] Each vehicle dimension in the vehicle data that corresponds to the target service is determined as the service input dimension corresponding to the vehicle data.

[0061] For example, the dimensions corresponding to the target service can be preset. Then, when obtaining vehicle data, the vehicle data can be analyzed to obtain the dimensions corresponding to the vehicle data. The method for this has been detailed above and will not be repeated here. Furthermore, after determining the dimensions corresponding to the vehicle data, the dimensions that match the preset dimensions of the target service can be determined as the service input dimensions corresponding to the vehicle data.

[0062] For example, for target service A, its pre-set multiple dimensions include A1-A4. The dimensions corresponding to the vehicle data determined based on vehicle data include A1, A2, A3, and B1. In this scenario, A1, A2, and A3 can be determined as the service input dimensions corresponding to the vehicle data. This ensures the diversity of dimensions while improving the correlation between the service input dimensions and the target service, thereby improving the accuracy of the training samples.

[0063] Based on the historical operation data, the operation dimension of the associated object corresponding to the target service in the vehicle, as well as the time dimension and frequency dimension of the operation on the target service, are determined as the service input dimension corresponding to the historical operation data.

[0064] The associated objects corresponding to the target service can be various operable objects in the vehicle, or the associated objects corresponding to the target service can be set in advance according to the actual application scenario. This disclosure does not limit this.

[0065] As mentioned above, historical operation data can be analyzed to identify operation records for various objects within the vehicle. For example, if the target service is air conditioning temperature control, the associated objects corresponding to this target service could be the air conditioner, seats, windows, etc., as described above. In this embodiment, after identifying the operation records for associated objects, these operation records can be used as the aforementioned operation dimensions. The operation dimensions of associated objects can reflect user operating habits to some extent. For instance, a user operating seat heating or air conditioning heating suggests the user is sensitive to temperature, while a user opening windows to a greater extent suggests the user prefers natural ventilation. Therefore, user operating habits in the vehicle can be modeled, ensuring the accuracy and comprehensiveness of the input features in the training samples.

[0066] Furthermore, based on historical operation data, the time and frequency of user operations on the target service can be determined. For example, the time when the user opens the target service can be statistically analyzed on a daily basis, and the frequency of the user opening the target service within the corresponding time period of the historical operation record can be statistically analyzed to further analyze the user's operation habits and provide effective data support for subsequent training of the service model.

[0067] Therefore, the above technical solution can filter the features in the vehicle that are related to the user's activation of the target service by analyzing vehicle data and historical operation data. This ensures the diversity of the input features of the service determination model while maintaining the correlation between the dimensions of the input features and the target service. This can reduce the amount of data required to train the service determination model to a certain extent, while improving the training efficiency and accuracy of the service determination model.

[0068] In step 22, the various service input dimensions are concatenated to obtain multiple candidate combination dimensions, wherein each candidate combination dimension contains at least two service input dimensions. For example, as described above, if multiple service input dimensions such as temperature, humidity, time, frequency, window operation, and seat operation are determined, then for a two-dimensional candidate combination dimension, any two dimensions can be selected and combined to obtain candidate combination dimensions, such as the candidate combination dimension corresponding to time-temperature, the candidate combination dimension corresponding to humidity-time, the combination dimension corresponding to temperature-window operation, etc., as well as candidate combination dimensions containing more dimensions, such as temperature-humidity-time. The determination method for other candidate combination dimensions is similar and will not be elaborated here.

[0069] In step 23, features under the candidate combination dimension are extracted based on vehicle data and historical operation data to obtain the combination dimension features corresponding to the candidate combination dimension.

[0070] Once the candidate combination dimension is determined, the corresponding value data can be obtained so that the service determination model can be trained based on the value data.

[0071] In one possible embodiment, an exemplary implementation of extracting features from the candidate combination dimension based on the vehicle data and the historical operation data to obtain the combined dimension features corresponding to the candidate combination dimension is as follows, which may include:

[0072] The dimension data for each dimension in the candidate combination dimensions is determined from the vehicle data and the historical operation data.

[0073] For example, if the candidate combination dimension is the combination dimension corresponding to time and temperature, then the dimension data under the temperature dimension can be obtained based on vehicle data, and the dimension data under the time dimension can be determined based on historical operation data.

[0074] Then, for each candidate combination dimension, the target dimension data under each dimension of the candidate combination dimension are concatenated to obtain the combination dimension feature corresponding to the candidate combination dimension.

[0075] Wherein, if the dimensional data under the dimension is numerical data, then the target dimensional data under the dimension is the discrete dimensional data under the dimension obtained after bucketing the dimensional data under the dimension; if the dimensional data under the dimension is non-numerical data, then the target dimensional data under the dimension is the dimensional data under the dimension.

[0076] As in the example above, if the data in the time dimension is numerical, then the data in this dimension can be pre-divided into buckets. For example, the data in the time dimension can be divided into buckets according to preset time periods. For example, time can be divided into four time periods: morning (e.g., 7:00-12:00), afternoon (e.g., 12:00-18:00), evening (e.g., 18:00-23:00), and night (e.g., 23:00-7:00). This bucketing process will obtain the corresponding discrete dimension data, that is, the numerical data in the time dimension will be transformed into discrete dimension data with values ​​in the morning, afternoon, evening, and night, which is the target dimension data in the time dimension.

[0077] Similarly, for dimensional data under the temperature dimension, it can be binned according to the temperature range. For example, the discrete dimensional data corresponding to temperature data above the upper limit of the temperature range is high temperature, the discrete dimensional data corresponding to temperature data below the lower limit of the temperature range is low temperature, and the discrete dimensional data corresponding to temperature data within the temperature range is comfort temperature. Thus, the dimensional data under the temperature dimension can be transformed into discrete dimensional data with values ​​such as high temperature, comfort temperature, and low temperature.

[0078] The binning of different numerical data can be performed according to their corresponding dimensions, and this disclosure does not impose any limitations on this. For non-numerical data, binning is not required, and its corresponding dimension data can be used as the target dimension data. For example, the operation dimension of a related object, such as the operation dimension of a car window, can have the corresponding dimension data as "open" and "close," and can be directly used as the target dimension data.

[0079] In this embodiment, for each candidate combination dimension, the target dimension data under each dimension of the candidate combination dimension are concatenated to obtain the combination dimension feature corresponding to the candidate combination dimension. For example, if the candidate combination dimension includes a time dimension and a temperature dimension, then each target dimension data under the time dimension can be concatenated with each target dimension data under the temperature dimension, resulting in the following combination dimension features: morning-high temperature, afternoon-high temperature, evening-high temperature, night-high temperature, morning-comfortable temperature, afternoon-comfortable temperature, evening-comfortable temperature, night-comfortable temperature, morning-low temperature, afternoon-low temperature, evening-low temperature, night-low temperature.

[0080] Therefore, by using the above technical solution to bin the numerical data, the numerical data can be discretized. This avoids the situation where too many combined dimensional features lead to an excessive amount of training sample data when combining dimensions. At the same time, discretization can be used to model user habits, making it easier to obtain user operation habits and improve the prediction accuracy of the service determination model.

[0081] In step 24, the target combination dimension is determined from the candidate combination dimensions based on the combination dimension features and vehicle data.

[0082] For example, taking air conditioning temperature control as an example, in the time dimension, the probability of activating the target service in the morning, afternoon, evening, and night is 45%, 45%, 5%, and 5%, respectively. In the temperature dimension, the probability of activating the target service in the afternoon corresponding to high temperature, comfortable temperature, and low temperature is 45%, 10%, and 45%, respectively. Therefore, in a single dimension, the probability of activating the target service in the morning and afternoon is similar, and the probability of activating the target service in high temperature and low temperature is similar. In this case, it is difficult to obtain accurate results when recommending target services based on features under a single dimension. In this embodiment, candidate combined dimensions are filtered by combining dimensional features.

[0083] For example, the step of determining the target combined dimension among the candidate combined dimensions based on the combined dimension features and the historical operation data may include:

[0084] Based on the historical operation data, determine the service results corresponding to each of the combined dimension features under the candidate combined dimension.

[0085] For example, the service results corresponding to each combined dimension feature can be determined based on historical operation data regarding operations on the target service. For ease of description, this can be simplified as follows: the time dimension includes morning and afternoon, and the temperature dimension includes high temperature and low temperature. Therefore, the combined dimension features include morning-high temperature, morning-low temperature, afternoon-high temperature, and afternoon-low temperature. In this embodiment, the service results under the features of morning-high temperature, morning-low temperature, afternoon-high temperature, and afternoon-low temperature, i.e., the probability of the target service being activated, can be determined based on historical operation data. For example, the probabilities of the target service being activated under the features of morning-high temperature, morning-low temperature, afternoon-high temperature, and afternoon-low temperature are 75%, 15%, 75%, and 5%, respectively.

[0086] If the difference in the proportion of confirmed services indicated by the service results corresponding to two combined dimension features under the candidate combined dimension exceeds a preset threshold, then the candidate combined dimension is determined as the target combined dimension.

[0087] For example, the preset threshold can be set according to the actual application scenario. For example, if it is set to 30%, as shown above, the difference in the proportion of confirmed service indicators for the morning-high temperature and morning-low temperature services under the candidate combination dimension is 60%, which exceeds the preset threshold of 30%. This means that different values ​​under the candidate combination dimension have a significant impact on the service results of the target service. That is, for the target service, the candidate combination dimension is a dimension that has a significant impact on it. At this time, the candidate combination dimension can be determined as the target combination dimension.

[0088] Therefore, combinations can be made based on a single dimension, and the combined dimensions that have an impact on the determination of the target service can be identified. This improves the feature mining depth in the training samples, further enhances the diversity and comprehensiveness of the input features in the service determination model, and provides effective data support for training the service determination model.

[0089] In step 25, the combined dimension of each service input dimension and the target dimension is determined as the target dimension.

[0090] Therefore, the target dimension corresponding to the determined service training sample can include multiple dimensions determined by vehicle data and historical operation data. At the same time, it can also explore the impact of the combination of multiple dimensions on the target service, thereby ensuring the feature matching degree between the target dimension and the determined target service and improving the training efficiency and accuracy of the service determination model.

[0091] In one possible embodiment, prior to the step of extracting feature dimensions based on the vehicle data and the historical operation data, the method may further include:

[0092] The vehicle data and the historical operation data are preprocessed to obtain processed vehicle data and processed historical operation data. The data preprocessing includes at least one of data format checking, default value filling, and numerical data binning.

[0093] The step of extracting feature dimensions based on the vehicle data and the historical operation data to obtain the target dimension corresponding to the input features of the service training sample includes:

[0094] Feature dimensions are extracted based on the processed vehicle data and the processed historical operation data to obtain the target dimension corresponding to the input features of the service training sample.

[0095] For example, to ensure consistent data processing, preprocessing is performed on the acquired vehicle data and historical operation data. This includes data format checks to verify the validity of the acquired data format, such as determining whether temperature and humidity data are numerical data. Another example is window operation data during driving; if no corresponding operation data is detected, it can be filled with default values, which can be disabled to ensure consistent labeling of input data in subsequent service training samples. Furthermore, numerical data can be binned; the implementation of binning has been detailed above and will not be repeated here.

[0096] Therefore, the above technical solution can preprocess the acquired data to bring it to the same standard, which facilitates subsequent unified processing of the data, improves the generation efficiency of service training samples and the uniformity of service training samples, and avoids the impact of formal errors of service training samples on the training efficiency of service determination models.

[0097] In one possible embodiment, an exemplary implementation of extracting features from the vehicle data and the historical operation data according to the target dimension to obtain service input features and service results in the service training samples is as follows, which may include:

[0098] If there is first vehicle data indicating that the user should start the target service or receiving confirmation from the user in response to the output of the target service, then according to the target dimension, the dimension parameters under the target dimension are extracted from the first vehicle data and the historical operation data as the service input features, and the service result is determined to be starting the target service, so as to obtain the service training samples under positive samples.

[0099] For example, a user can actively activate a target service. For instance, if a user feels the vehicle's interior temperature is too high while driving, they can directly control the air conditioning to turn on cooling mode via a control button. The initial vehicle data generated based on this user action can then indicate that the user has actively activated the target service.

[0100] Furthermore, the dimensional parameters of each target dimension can be extracted from the first vehicle data (i.e., the vehicle data obtained at the moment when the user operates to start the target service) and historical operation data as service input features of the service training sample. Based on the user's operation confirming that the user turns on the air conditioner at this time, the service result of the service training sample can be determined as the start of the target service and used as a positive sample.

[0101] As another example, the user's activation of a target service can be achieved by the vehicle's control system recommending the target service to the user, who then confirms activation. For instance, if the vehicle recommends the target service to the user based on current vehicle information while it is in motion, and the user decides that the recommended target service can be activated, they can confirm the activation via voice reply or by selecting a confirmation button on the display interface. The first vehicle data generated at this time can also indicate that the user actively activates the target service. Service training samples can then be generated based on historical operation data and the first vehicle data, with the service result being the activation of the target service, which is then used as a positive sample.

[0102] Specifically, a service training sample can be generated based on the vehicle data corresponding to each user's operation on the target service and the historical operation data. For example, for each service training sample, the dimension parameters corresponding to each dimension in the target dimension can be extracted from the data corresponding to the service training sample, such as the temperature value under the temperature dimension, the humidity value under the humidity dimension, etc.

[0103] If an indication is received that the target service output has received a user's rejection operation, then based on the second vehicle data of the target dimension, the dimension parameters under the target dimension are extracted from the second vehicle data and the historical operation data as the service input features, and the target service result is determined to be a rejection of the target service, so as to obtain the service training samples under negative samples.

[0104] As an example, if a vehicle recommends a specific service to a user based on current vehicle information while the vehicle is in motion, and the user decides that the recommended service is unnecessary, they can refuse to activate the service through voice prompts or by selecting a "decline" button on the display interface. The generated second vehicle data can then instruct the user to refuse the service. Service training samples can be generated based on historical operation data and the second vehicle data, with the corresponding service result being "refuse the target service," and these samples are used as negative samples.

[0105] If the vehicle data indicates that the target service was not activated during the vehicle's operation, then the dimension parameters under the target dimension are extracted from the vehicle data and the historical operation data as the service input features, and the target service result is determined to be a rejection of the target service, so as to obtain the service training samples under negative samples.

[0106] In this embodiment, if the target service is not activated during vehicle operation, the data during the operation can be considered as negative sample data corresponding to the target service. That is, N records can be randomly selected from the vehicle data and historical operation data obtained during vehicle operation, and the dimension parameters under the target dimension can be extracted from each record. The service result is determined as rejecting the target service and is used as a negative sample.

[0107] Therefore, through the above technical solution, positive samples for the service determination model can be generated based on the user's active activation of the target service and the user's confirmation of the recommended target service. Conversely, negative samples can be generated based on the user's rejection of the recommended target service and the data from the driving process of activating the target service. This improves the matching degree between positive and negative samples and the user's actual driving process, thereby enhancing the prediction accuracy of the service determination model. Furthermore, it can also improve the diversity and generation efficiency of the training samples for the service determination model to some extent.

[0108] This disclosure also provides a method for determining vehicle services, wherein the method includes:

[0109] The system acquires target vehicle data and historical operation data provided by the target user corresponding to the target vehicle. The historical operation data provided by the target user can be determined based on the user's historical operation data recorded in the target vehicle. For example, the user driving the vehicle can be assumed to be the same user, i.e., the target user. Alternatively, the target user can be a user who logged into the vehicle management system before driving the vehicle; that is, the user corresponding to that login account is considered the target user. This user can provide their corresponding historical operation data through their login account, thus representing the operating habits of the current driver.

[0110] Subsequently, based on the target dimension in the service determination model corresponding to the target service, features are extracted from the target vehicle data and the historical operation data provided by the target user to obtain the current service features of the target vehicle. The service determination model for the target service is obtained based on the training method described above. The method for feature extraction based on the target dimension has been detailed above and will not be repeated here.

[0111] Based on the current service characteristics and the service determination model, determine whether the target service is a recommended service output by the target vehicle.

[0112] The current service features can be input into the service determination model to obtain the prediction result obtained by the service determination model. If the prediction result is to enable the target service, the target service can be determined as the recommended service output by the target vehicle, that is, the target service is recommended to the user. If the prediction result is to reject the target service, the target service will not be used as the recommended service output by the vehicle, that is, the target service will not be recommended at this time.

[0113] Therefore, by employing the aforementioned technical solution, recommended services can be determined for the user based on the current vehicle data and historical operation data provided by the target user. This allows for personalized recommendations tailored to different users' operating habits, improving the accuracy and effectiveness of vehicle service determination, while also enhancing the match between the determined services and the current driver, thus improving the user experience.

[0114] This disclosure also provides a training apparatus for a vehicle service determination model, such as... Figure 3 As shown, the device 10 includes:

[0115] The first acquisition module 100 is used to acquire vehicle data corresponding to the vehicle and historical operation data provided by the user.

[0116] The first processing module 200 is used to extract feature dimensions based on the vehicle data and the historical operation data to obtain the target dimension corresponding to the input features of the service training sample.

[0117] The first extraction module 300 is used to extract features from the vehicle data and the historical operation data according to the target dimension to obtain service input features and service results in the service training sample, wherein the service results include enabling the target service and rejecting the target service;

[0118] The training module 400 is used to train the service determination model by taking the service input features as input to the service determination model, taking the service results corresponding to the service input features as the target output of the service determination model, and determining the trained service determination model as the service determination model corresponding to the target service.

[0119] Optionally, the first processing module includes:

[0120] The first determining submodule is used to determine the various service input dimensions corresponding to the vehicle data and the historical operation data;

[0121] The first splicing submodule is used to splice the various service input dimensions to obtain multiple candidate combination dimensions, wherein the candidate combination dimensions include at least two service input dimensions;

[0122] The extraction submodule is used to extract features under the candidate combination dimension based on the vehicle data and the historical operation data, and obtain the combination dimension features corresponding to the candidate combination dimension.

[0123] The second determining submodule is used to determine the target combined dimension among the candidate combined dimensions based on the combined dimension features and the vehicle data.

[0124] The third determining submodule is used to determine each of the service input dimensions and the target combination dimension as the target dimension.

[0125] Optionally, the first determining submodule includes:

[0126] The fourth determining submodule is used to determine each vehicle dimension in the vehicle data that corresponds to the target service, as the service input dimension corresponding to the vehicle data;

[0127] The fifth determining submodule is used to determine, based on the historical operation data, the operation dimension of the associated object in the vehicle corresponding to the target service, as well as the time dimension and frequency dimension of the operation on the target service, as the service input dimension corresponding to the historical operation data.

[0128] Optionally, the extraction submodule includes:

[0129] The sixth determining submodule is used to determine the dimension data for each dimension in the candidate combination dimensions from the vehicle data and the historical operation data;

[0130] The second splicing submodule is used to splice the target dimension data under each dimension of the candidate combination dimension for each candidate combination dimension to obtain the combination dimension feature corresponding to the candidate combination dimension.

[0131] Wherein, if the dimensional data under the dimension is numerical data, then the target dimensional data under the dimension is the discrete dimensional data under the dimension obtained after bucketing the dimensional data under the dimension; if the dimensional data under the dimension is non-numerical data, then the target dimensional data under the dimension is the dimensional data under the dimension.

[0132] Optionally, the second determining submodule includes:

[0133] The seventh determining submodule is used to determine the service results corresponding to each of the combined dimension features under the candidate combined dimension based on the vehicle data;

[0134] The eighth determining submodule is used to determine the candidate combination dimension as the target combination dimension if the difference between the proportions of the service results indicated by the two combination dimension features under the candidate combination dimension exceeds a preset threshold.

[0135] Optionally, the device further includes:

[0136] The second processing module is used to perform data preprocessing on the vehicle data and the historical operation data before the first processing module extracts feature dimensions based on the vehicle data and the historical operation data, so as to obtain processed vehicle data and processed historical operation data. The data preprocessing includes at least one of data format checking, default value filling, and numerical data binning.

[0137] The first processing module is used for:

[0138] Feature dimensions are extracted based on the processed vehicle data and the processed historical operation data to obtain the target dimension corresponding to the input features of the service training sample.

[0139] Optionally, the first extraction module includes:

[0140] The ninth determining submodule is used to extract dimension parameters under the target dimension from the first vehicle data and the historical operation data according to the target dimension if there is first vehicle data that instructs the user to start the target service or receives the user's confirmation operation in response to the output of the target service, and to use them as the service input features, and determine the service result as starting the target service, so as to obtain the service training samples under positive samples.

[0141] The tenth determination submodule is used to extract the dimension parameters under the target dimension from the second vehicle data and the historical operation data according to the second vehicle data of the target dimension, as the service input features, and determine the target service result as the rejection target service if there is an indication that the output of the target service has received a user rejection operation, so as to obtain the service training sample under the negative sample.

[0142] The eleventh determination submodule is used to extract the dimension parameters under the target dimension from the vehicle data and the historical operation data according to the target dimension if the vehicle data indicates that the target service was not activated during the vehicle's driving process, and use them as the service input features, and determine that the target service result is to reject the target service, so as to obtain the service training samples under negative samples.

[0143] This disclosure also provides a vehicle service determination device, the device comprising:

[0144] The second acquisition module is used to acquire target vehicle data corresponding to the target vehicle and historical operation data provided by the target user corresponding to the target vehicle.

[0145] The second extraction module is used to extract features from the target vehicle data and the historical operation data provided by the target user based on the target dimension in the service determination model corresponding to the target service, and obtain the current service features of the target vehicle. The service determination model of the target service is obtained by any of the service determination models described above.

[0146] The determination module is used to determine whether the target service is a recommended service output by the target vehicle based on the current service characteristics and the service determination model.

[0147] The following is for reference. Figure 4 This diagram illustrates a structural schematic of an electronic device 600 suitable for implementing embodiments of the present disclosure. The terminal devices in the embodiments of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 4 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.

[0148] like Figure 4 As shown, electronic device 600 may include a processing device (e.g., a central processing unit, a graphics processor, etc.) 601, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 602 or a program loaded from storage device 608 into random access memory (RAM) 603. RAM 603 also stores various programs and data required for the operation of electronic device 600. Processing device 601, ROM 602, and RAM 603 are interconnected via bus 604. Input / output (I / O) interface 605 is also connected to bus 604.

[0149] Typically, the following devices can be connected to I / O interface 605: input devices 606 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 607 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 608 including, for example, magnetic tapes, hard disks, etc.; and communication devices 609. Communication device 609 allows electronic device 600 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 4 An electronic device 600 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively.

[0150] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 609, or installed from a storage device 608, or installed from a ROM 602. When the computer program is executed by the processing device 601, it performs the functions defined in the methods of embodiments of this disclosure.

[0151] It should be noted that the computer-readable medium described in this disclosure can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this disclosure, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this disclosure, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0152] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0153] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device.

[0154] The aforementioned computer-readable medium carries one or more programs. When the electronic device executes the aforementioned one or more programs, the electronic device causes the following actions: to acquire vehicle data corresponding to the vehicle and historical operation data provided by the user; to extract feature dimensions based on the vehicle data and the historical operation data to obtain a target dimension corresponding to the input features of the service training sample; to extract features from the vehicle data and the historical operation data based on the target dimension to obtain service input features and service results in the service training sample, wherein the service results include enabling the target service and rejecting the target service; to train the service determination model by using the service input features as input and the service results corresponding to the service input features as the target output of the service determination model; and to determine the trained service determination model as the service determination model corresponding to the target service.

[0155] Alternatively, the aforementioned computer-readable medium carries one or more programs that, when executed by the electronic device, cause the electronic device to: acquire target vehicle data corresponding to the target vehicle and historical operation data provided by the target user corresponding to the target vehicle; extract features from the target vehicle data and the historical operation data provided by the target user based on the target dimension in the service determination model corresponding to the target service to obtain the current service features of the target vehicle, wherein the service determination model of the target service is obtained based on the training method of any of the service determination models described above; and determine whether the target service is a recommended service output by the target vehicle based on the current service features and the service determination model.

[0156] Computer program code for performing the operations of this disclosure can be written in one or more programming languages ​​or a combination thereof, including but not limited to object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0157] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0158] The modules described in the embodiments of this disclosure can be implemented in software or hardware. The names of the modules are not necessarily limiting in certain circumstances; for example, the first acquisition module can also be described as "a module that acquires vehicle data corresponding to the vehicle and historical operation data provided by the user."

[0159] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0160] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0161] According to one or more embodiments of this disclosure, Example 1 provides a method for training a vehicle service determination model, wherein the method includes:

[0162] Obtain vehicle data corresponding to the vehicle and historical operation data provided by the user;

[0163] Based on the vehicle data and the historical operation data, feature dimensions are extracted to obtain the target dimension corresponding to the input features of the service training sample;

[0164] Based on the target dimension, features are extracted from the vehicle data and the historical operation data to obtain service input features and service results in the service training sample, wherein the service results include enabling the target service and rejecting the target service;

[0165] The service input features are used as input to the service determination model, and the service results corresponding to the service input features are used as the target output of the service determination model. The service determination model is trained, and the trained service determination model is determined as the service determination model corresponding to the target service.

[0166] According to one or more embodiments of this disclosure, Example 2 provides the method of Example 1, wherein the step of extracting feature dimensions based on the vehicle data and the historical operation data to obtain the target dimension corresponding to the input features of the service training sample includes:

[0167] Determine the various service input dimensions corresponding to the vehicle data and the historical operation data;

[0168] Each of the service input dimensions is concatenated to obtain multiple candidate combination dimensions, wherein each candidate combination dimension contains at least two service input dimensions;

[0169] Based on the vehicle data and the historical operation data, features under the candidate combination dimension are extracted to obtain the combination dimension features corresponding to the candidate combination dimension;

[0170] Based on the combined dimension features and the vehicle data, determine the target combined dimension among the candidate combined dimensions;

[0171] Each of the service input dimensions and the target combination dimension is determined as the target dimension.

[0172] According to one or more embodiments of this disclosure, Example 3 provides the method of Example 2, wherein determining the various service input dimensions corresponding to the vehicle data and the historical operation data includes:

[0173] Determine each vehicle dimension in the vehicle data that corresponds to the target service, and use it as the service input dimension corresponding to the vehicle data;

[0174] Based on the historical operation data, the operation dimension of the associated object corresponding to the target service in the vehicle, as well as the time dimension and frequency dimension of the operation on the target service, are determined as the service input dimension corresponding to the historical operation data.

[0175] According to one or more embodiments of this disclosure, Example 4 provides the method of Example 2, wherein the step of extracting features under the candidate combination dimension based on the vehicle data and the historical operation data to obtain the combination dimension features corresponding to the candidate combination dimension includes:

[0176] Dimension data for each dimension in the candidate combination dimensions are determined from the vehicle data and the historical operation data;

[0177] For each candidate combination dimension, the target dimension data under each dimension of the candidate combination dimension are concatenated to obtain the combination dimension feature corresponding to the candidate combination dimension.

[0178] Wherein, if the dimensional data under the dimension is numerical data, then the target dimensional data under the dimension is the discrete dimensional data under the dimension obtained after bucketing the dimensional data under the dimension; if the dimensional data under the dimension is non-numerical data, then the target dimensional data under the dimension is the dimensional data under the dimension.

[0179] According to one or more embodiments of this disclosure, Example 5 provides the method of Example 2, wherein determining the target combined dimension among the candidate combined dimensions based on the combined dimension features and the vehicle data includes:

[0180] Based on the vehicle data, determine the service results corresponding to each feature of the combined dimension under the candidate combined dimension;

[0181] If the difference between the proportions of the enabled target service indicated by the service results corresponding to two combined dimension features under the candidate combined dimension exceeds a preset threshold, then the candidate combined dimension is determined as the target combined dimension.

[0182] According to one or more embodiments of this disclosure, Example 6 provides the method of Example 1, wherein, prior to the step of extracting feature dimensions based on the vehicle data and the historical operation data, the method further includes:

[0183] The vehicle data and the historical operation data are preprocessed to obtain processed vehicle data and processed historical operation data. The data preprocessing includes at least one of data format checking, default value filling, and numerical data binning.

[0184] The step of extracting feature dimensions based on the vehicle data and the historical operation data to obtain the target dimension corresponding to the input features of the service training sample includes:

[0185] Feature dimensions are extracted based on the processed vehicle data and the processed historical operation data to obtain the target dimension corresponding to the input features of the service training sample.

[0186] According to one or more embodiments of this disclosure, Example 7 provides the method of Example 1, wherein the step of extracting features from the vehicle data and the historical operation data according to the target dimension to obtain service input features and service results in the service training sample includes:

[0187] If there is first vehicle data indicating that the user should start the target service or receiving confirmation from the user in response to the output of the target service, then according to the target dimension, the dimension parameters under the target dimension are extracted from the first vehicle data and the historical operation data as the service input features, and the service result is determined to be starting the target service, so as to obtain the service training samples under positive samples.

[0188] If there is an indication that the output of the target service has received a user's rejection operation, then based on the second vehicle data of the target dimension, the dimension parameters under the target dimension are extracted from the second vehicle data and the historical operation data as the service input features, and the target service result is determined to be a rejection of the target service, so as to obtain the service training samples under negative samples.

[0189] If the vehicle data indicates that the target service was not activated during the vehicle's operation, then based on the target dimension, the dimension parameters under the target dimension are extracted from the vehicle data and the historical operation data as the service input features, and the target service result is determined to be a rejection of the target service, so as to obtain the service training samples under negative samples.

[0190] According to one or more embodiments of this disclosure, Example 8 provides a vehicle service determination method, the method comprising:

[0191] Obtain the target vehicle data corresponding to the target vehicle and the historical operation data provided by the target user corresponding to the target vehicle;

[0192] Based on the target dimension in the service determination model corresponding to the target service, feature extraction is performed from the target vehicle data and the historical operation data provided by the target user to obtain the current service features of the target vehicle. The service determination model of the target service is obtained based on the training method of any of the service determination models described in Examples 1-7.

[0193] Based on the current service characteristics and the service determination model, determine whether the target service is a recommended service output by the target vehicle.

[0194] According to one or more embodiments of this disclosure, Example 9 provides a training apparatus for a vehicle service determination model, the apparatus comprising:

[0195] The first acquisition module is used to acquire vehicle data corresponding to the vehicle and historical operation data provided by the user.

[0196] The first processing module is used to extract feature dimensions based on the vehicle data and the historical operation data to obtain the target dimension corresponding to the input features of the service training sample.

[0197] The first extraction module is used to extract features from the vehicle data and the historical operation data according to the target dimension to obtain service input features and service results in the service training sample, wherein the service results include enabling the target service and rejecting the target service;

[0198] The training module is used to train the service determination model by taking the service input features as input, the service results corresponding to the service input features as the target output of the service determination model, and determining the trained service determination model as the service determination model corresponding to the target service.

[0199] According to one or more embodiments of this disclosure, Example 10 provides a vehicle service determination apparatus, wherein the apparatus includes:

[0200] The second acquisition module is used to acquire target vehicle data corresponding to the target vehicle and historical operation data provided by the target user corresponding to the target vehicle.

[0201] The second extraction module is used to extract features from the target vehicle data and the historical operation data provided by the target user based on the target dimension in the service determination model corresponding to the target service, and obtain the current service features of the target vehicle. The service determination model of the target service is obtained based on the training method of any of the service determination models described in Examples 1-7.

[0202] The determination module is used to determine whether the target service is a recommended service output by the target vehicle based on the current service characteristics and the service determination model.

[0203] According to one or more embodiments of the present disclosure, Example 11 provides a computer-readable medium having a computer program stored thereon that, when executed by a processing device, implements the steps of the method described in any one of Examples 1-8.

[0204] According to one or more embodiments of this disclosure, Example 12 provides an electronic device, including:

[0205] A storage device on which computer programs are stored;

[0206] A processing device for executing the computer program in the storage device to implement the steps of any one of the methods in Examples 1-8.

[0207] The above description is merely a preferred embodiment of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features disclosed in this disclosure that have similar functions.

[0208] Furthermore, while the operations are described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or in a sequential order. In certain environments, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of this disclosure. Certain features described in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments.

[0209] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative forms of implementing the claims. Regarding the apparatus in the above embodiments, the specific manner in which the various modules perform their operations has been described in detail in the embodiments relating to the method, and will not be elaborated upon here.

Claims

1. A training method for a vehicle service determination model, characterized in that, The method includes: Obtain vehicle data corresponding to the vehicle and historical operation data provided by the user; Based on the vehicle data and the historical operation data, feature dimensions are extracted to obtain the target dimension corresponding to the input features of the service training sample; Based on the target dimension, features are extracted from the vehicle data and the historical operation data to obtain service input features and service results in the service training sample, wherein the service results include enabling the target service and rejecting the target service; The service input features are used as the input to the service determination model, and the service results corresponding to the service input features are used as the target output of the service determination model. The service determination model is trained, and the trained service determination model is determined as the service determination model corresponding to the target service. The step of extracting feature dimensions based on the vehicle data and the historical operation data to obtain the target dimension corresponding to the input features of the service training sample includes: Determine the various service input dimensions corresponding to the vehicle data and the historical operation data; Each of the service input dimensions is concatenated to obtain multiple candidate combination dimensions, wherein each candidate combination dimension contains at least two service input dimensions; Based on the vehicle data and the historical operation data, features under the candidate combination dimension are extracted to obtain the combination dimension features corresponding to the candidate combination dimension; Based on the combined dimension features and the vehicle data, determine the target combined dimension among the candidate combined dimensions; Each of the service input dimensions and the target combination dimension is determined as the target dimension.

2. The method according to claim 1, characterized in that, The determination of the various service input dimensions corresponding to the vehicle data and the historical operation data includes: Determine each vehicle dimension in the vehicle data that corresponds to the target service, and use it as the service input dimension corresponding to the vehicle data; Based on the historical operation data, the operation dimension of the associated object corresponding to the target service in the vehicle, as well as the time dimension and frequency dimension of the operation on the target service, are determined as the service input dimension corresponding to the historical operation data.

3. The method according to claim 1, characterized in that, The step of extracting features under the candidate combination dimension based on the vehicle data and the historical operation data to obtain the combination dimension features corresponding to the candidate combination dimension includes: Dimension data for each dimension in the candidate combination dimensions are determined from the vehicle data and the historical operation data; For each candidate combination dimension, the target dimension data under each dimension of the candidate combination dimension are concatenated to obtain the combination dimension feature corresponding to the candidate combination dimension. Wherein, if the dimensional data under the dimension is numerical data, then the target dimensional data under the dimension is the discrete dimensional data under the dimension obtained after bucketing the dimensional data under the dimension; if the dimensional data under the dimension is non-numerical data, then the target dimensional data under the dimension is the dimensional data under the dimension.

4. The method according to claim 1, characterized in that, The step of determining the target combined dimension among the candidate combined dimensions based on the combined dimension features and the vehicle data includes: Based on the vehicle data, determine the service results corresponding to each feature of the combined dimension under the candidate combined dimension; If the difference between the proportions of the enabled target service indicated by the service results corresponding to two combined dimension features under the candidate combined dimension exceeds a preset threshold, then the candidate combined dimension is determined as the target combined dimension.

5. The method according to claim 1, characterized in that, Before the step of extracting feature dimensions based on the vehicle data and the historical operation data, the method further includes: The vehicle data and the historical operation data are preprocessed to obtain processed vehicle data and processed historical operation data. The data preprocessing includes at least one of data format checking, default value filling, and numerical data binning. The step of extracting feature dimensions based on the vehicle data and the historical operation data to obtain the target dimension corresponding to the input features of the service training sample includes: Feature dimensions are extracted based on the processed vehicle data and the processed historical operation data to obtain the target dimension corresponding to the input features of the service training sample.

6. The method according to claim 1, characterized in that, The step of extracting features from the vehicle data and the historical operation data according to the target dimension to obtain the service input features and service results in the service training samples includes: If there is first vehicle data indicating that the user should start the target service or receiving confirmation from the user in response to the output of the target service, then according to the target dimension, the dimension parameters under the target dimension are extracted from the first vehicle data and the historical operation data as the service input features, and the service result is determined to be starting the target service, so as to obtain the service training samples under positive samples. If there is an indication that the output of the target service has received a user's rejection operation, then based on the second vehicle data of the target dimension, the dimension parameters under the target dimension are extracted from the second vehicle data and the historical operation data as the service input features, and the target service result is determined to be a rejection of the target service, so as to obtain the service training samples under negative samples. If the vehicle data indicates that the target service was not activated during the vehicle's operation, then based on the target dimension, the dimension parameters under the target dimension are extracted from the vehicle data and the historical operation data as the service input features, and the target service result is determined to be a rejection of the target service, so as to obtain the service training samples under negative samples.

7. A method for determining vehicle service, characterized in that, The method includes: Obtain the target vehicle data corresponding to the target vehicle and the historical operation data provided by the target user corresponding to the target vehicle; Based on the target dimension in the service determination model corresponding to the target service, feature extraction is performed from the target vehicle data and the historical operation data provided by the target user to obtain the current service features of the target vehicle. The service determination model of the target service is obtained based on the training method of the service determination model according to any one of claims 1-6. Based on the current service characteristics and the service determination model, determine whether the target service is a recommended service output by the target vehicle.

8. A training apparatus for a vehicle service determination model, characterized in that, The device includes: The first acquisition module is used to acquire vehicle data corresponding to the vehicle and historical operation data provided by the user. The first processing module is used to extract feature dimensions based on the vehicle data and the historical operation data to obtain the target dimension corresponding to the input features of the service training sample. The first extraction module is used to extract features from the vehicle data and the historical operation data according to the target dimension to obtain service input features and service results in the service training sample, wherein the service results include enabling the target service and rejecting the target service; The training module is used to take the service input features as input to the service determination model, take the service results corresponding to the service input features as the target output of the service determination model, train the service determination model, and determine the trained service determination model as the service determination model corresponding to the target service. The first processing module includes: The first determining submodule is used to determine the various service input dimensions corresponding to the vehicle data and the historical operation data; The first splicing submodule is used to splice the various service input dimensions to obtain multiple candidate combination dimensions, wherein the candidate combination dimensions include at least two service input dimensions; The extraction submodule is used to extract features under the candidate combination dimension based on the vehicle data and the historical operation data, and obtain the combination dimension features corresponding to the candidate combination dimension. The second determining submodule is used to determine the target combined dimension among the candidate combined dimensions based on the combined dimension features and the vehicle data. The third determining submodule is used to determine each of the service input dimensions and the target combination dimension as the target dimension.

9. A vehicle service determination device, characterized in that, The device includes: The second acquisition module is used to acquire target vehicle data corresponding to the target vehicle and historical operation data provided by the target user corresponding to the target vehicle. The second extraction module is used to extract features from the target vehicle data and the historical operation data provided by the target user based on the target dimension in the service determination model corresponding to the target service, and obtain the current service features of the target vehicle. The service determination model of the target service is obtained based on the training method of the service determination model according to any one of claims 1-6. The determination module is used to determine whether the target service is a recommended service output by the target vehicle based on the current service characteristics and the service determination model.

10. A computer-readable medium having a computer program stored thereon, characterized in that, When executed by the processing device, the program implements the steps of the method described in any one of claims 1-7.

11. An electronic device, characterized in that, include: A storage device on which computer programs are stored; A processing device for executing the computer program in the storage device to implement the steps of the method according to any one of claims 1-7.

Images