CN115470403A - Real-time updating method and device of vehicle service recommendation model, vehicle and medium - Google Patents
Real-time updating method and device of vehicle service recommendation model, vehicle and medium Download PDFInfo
- Publication number
- CN115470403A CN115470403A CN202211018341.2A CN202211018341A CN115470403A CN 115470403 A CN115470403 A CN 115470403A CN 202211018341 A CN202211018341 A CN 202211018341A CN 115470403 A CN115470403 A CN 115470403A
- Authority
- CN
- China
- Prior art keywords
- user
- data
- real
- initial
- recommendation model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application relates to the technical field of vehicle service, in particular to a real-time updating method, a real-time updating device, a real-time updating vehicle and a real-time updating medium of a vehicle service recommendation model, wherein the method comprises the following steps: a real-time updating method for a vehicle service recommendation model is characterized by comprising the following steps: acquiring real-time data of a user, and recommending vehicle service according to the real-time data of the user; obtaining user feedback data according to feedback recommended by vehicle service; reading an initial recommendation model at the current moment according to user feedback data; and updating the initial recommendation model according to the user feedback data to obtain a service recommendation model, thereby realizing updating of recommendation service according to the real-time requirements of the user while ensuring personalized recommendation of vehicle services to the user.
Description
Technical Field
The present application relates to the field of vehicle service technologies, and in particular, to a method, an apparatus, a vehicle, and a medium for updating a vehicle service recommendation model in real time.
Background
In the related art, a preset recommendation model can directly recommend personalized services for a user. Accordingly, however, accurate service recommendation cannot be provided according to the implementation requirements of the user by using the preset recommendation model. Therefore, how to update the recommended service according to the real-time requirement of the user while ensuring that the vehicle service is recommended to the user in a personalized manner becomes an urgent problem to be solved.
Disclosure of Invention
Aiming at solving one or more technical problems, the application provides a real-time updating method of a vehicle service recommendation model, which comprises the following steps:
acquiring user real-time data, and recommending vehicle service according to the user real-time data;
obtaining user feedback data according to the feedback recommended by the vehicle service;
reading an initial recommendation model at the current moment according to the user feedback data;
and updating the initial recommendation model according to the user feedback data to obtain a service recommendation model.
The real-time updating method provided by the application can update the initial recommendation model according to the feedback of the user to the vehicle service recommended by the current initial recommendation model, so that more accurate vehicle service recommendation can be provided according to the implementation requirements of the user. Specifically, the method comprises the steps of firstly obtaining real-time data of a user, analyzing the real-time data of the user to provide vehicle services which may be needed by the user at the current moment, and forming vehicle service recommendations. Further, the method can obtain user feedback data according to the feedback of the user to the vehicle service. Wherein the user feedback data may be the user's response to the vehicle service recommendation. Furthermore, the method can read the corresponding initial recommendation model at the moment of the user feedback data, and further update the initial recommendation model according to the user feedback data so as to obtain the updated service recommendation model according to the user feedback data. Through the service recommendation model, the method can divide the accuracy of the vehicle service recommended by the initial recommendation model, judge the application condition of the recommended service according to the user feedback data, further obtain the service recommendation model after the recommendation accuracy is improved, and update the recommended service according to the real-time requirement of the user while ensuring the personalized recommendation of the vehicle service to the user so as to improve the service recommendation accuracy.
In some embodiments, the obtaining real-time user data and recommending vehicle services according to the real-time user data includes:
monitoring vehicle environment data according to the user real-time data;
combining the user real-time data and the vehicle environment data to sort preset vehicle services to obtain sorting information;
and sequentially recommending the vehicle service to the preset vehicle service according to the sequencing information.
In some embodiments, the obtaining user feedback data based on the feedback recommended by the vehicle service includes:
acquiring first user feedback within a preset acquisition duration; the first user feedback comprises recommended services accepted by the user within the preset acquisition duration;
after the preset acquisition duration, acquiring second user feedback; the second user feedback comprises recommended services which are not responded or refused by the user within the preset acquisition duration;
continuously acquiring third user feedback initiated by the user; wherein the third user feedback comprises a service actively initiated by the user;
generating the user feedback data from the first user feedback, the second user feedback, and the third user feedback.
In certain embodiments, the method further comprises:
and if the user feedback data does not contain the second user feedback and the third user feedback, taking the initial recommendation model at the current moment as a service recommendation model.
In certain embodiments, the method further comprises:
generating an initial Q network according to a preset network structure;
and obtaining an initial data set, and performing model training on the initial Q network according to the initial data set to obtain the initial recommendation model.
In certain embodiments, the method further comprises:
updating the initial data set according to the user feedback data to obtain an updated data set;
and performing model training on the initial Q network through the updating data set according to a preset initialization duration so as to update the initial recommendation model.
In some embodiments, said updating said initial data set according to said user feedback data to obtain an updated data set comprises:
calling the real-time user data at the next moment according to the user feedback data;
generating first grading data according to feedback of the user on the vehicle service recommendation;
and updating the initial data set according to the first grading data, the vehicle service recommendation, the user real-time data at the current moment and the user real-time data at the next moment to obtain an updated data set.
The updating the initial recommendation model according to the user feedback data to obtain a service recommendation model comprises:
grading the user feedback data according to a preset grading rule to obtain second grading data;
and updating the model parameters of the initial recommendation model according to the user feedback data and the second grading data to obtain a service recommendation model.
The application provides a real-time update device of vehicle service recommendation model, includes:
the online updating module is used for acquiring real-time user data and recommending vehicle service according to the real-time user data; the vehicle service recommendation system is also used for obtaining user feedback data according to the feedback recommended by the vehicle service; the system is also used for reading the initial recommendation model at the current moment according to the user feedback data; and the initial recommendation model is updated according to the user feedback data to obtain a service recommendation model.
In certain embodiments, the apparatus further comprises:
the offline updating module is used for generating an initial Q network according to a preset network structure; and the method is also used for obtaining an initial data set and carrying out model training on the initial Q network according to the initial data set to obtain the initial recommendation model. And the initial recommendation model is updated according to the user feedback data to obtain an updated data set, so that the initial recommendation model is updated according to the updated data set.
The present application provides a vehicle, comprising: the vehicle service recommendation system comprises a memory, a processor and a program stored on the memory and capable of running on the processor, wherein the processor executes the program to realize the real-time updating method of the vehicle service recommendation model.
The present application provides a computer readable storage medium having stored thereon a computer program for execution by a processor for implementing a method of real-time updating of a vehicle service recommendation model as claimed in any one of the preceding claims.
Additional aspects and advantages of the present application will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the present application.
Drawings
The above and/or additional aspects and advantages of the present application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
fig. 1 is a schematic flowchart of a method for updating a vehicle service recommendation model in real time according to an embodiment of the present application.
Fig. 2 is a schematic structural diagram of an initial Q network provided according to an embodiment of the present application.
Fig. 3 is a schematic structural diagram of a device for updating a vehicle service recommendation model in real time according to an embodiment of the present application.
Fig. 4 is a schematic interaction diagram between an online update module and an offline update module provided according to an embodiment of the present application.
Fig. 5 is a schematic structural diagram of a vehicle according to an embodiment of the present application.
Description of reference numerals: a real-time update device 10, an online update module 11, an offline update module 12, a vehicle 100, a processor 110, and a memory 120.
Detailed Description
Reference will now be made in detail to the embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the drawings are exemplary and intended to be used for explaining the present application and should not be construed as limiting the present application.
Specifically, fig. 1 is a method for updating a vehicle service recommendation model in real time according to an embodiment of the present application, including:
s1, acquiring real-time data of a user, and recommending vehicle service according to the real-time data of the user;
s2, obtaining user feedback data according to feedback recommended by the vehicle service;
s3, reading an initial recommendation model at the current moment according to the user feedback data;
and S4, updating the initial recommendation model according to the user feedback data to obtain a service recommendation model.
The real-time updating method provided by the application can update the initial recommendation model according to the feedback of the user to the vehicle service recommended by the current initial recommendation model, so that more accurate vehicle service recommendation can be provided according to the implementation requirements of the user. Specifically, the method comprises the steps of firstly obtaining real-time data of a user, analyzing the real-time data of the user to provide vehicle services which may be needed by the user at the current moment, and forming vehicle service recommendations. Further, the method can obtain user feedback data according to the feedback of the user to the vehicle service. Wherein the user feedback data may be the user's response to the vehicle service recommendation. Furthermore, the method can read the corresponding initial recommendation model at the moment of the user feedback data, and further update the initial recommendation model according to the user feedback data so as to obtain the service recommendation model updated according to the user feedback data. By the service recommendation model, the accuracy of the vehicle service recommended by the initial recommendation model can be divided, the application condition of the recommended service by the user in the user feedback data is judged, the service recommendation model with the recommendation accuracy improved is further obtained, the vehicle service can be recommended in an individualized mode by the user, and meanwhile the recommendation service can be updated according to the real-time requirements of the user, so that the service recommendation accuracy is improved.
In certain embodiments, the step is a cycle period within the overall cycle body. The method can further obtain the real-time data of the user again and execute the subsequent steps after the service recommendation model is generated, and further takes the generated service recommendation model as an initial recommendation model to further update according to the feedback data of the user. So as to ensure the accuracy of service recommendation of the finally updated service recommendation model.
Specifically, the method can also provide vehicle service recommendation for the user through the initial recommendation model so as to realize targeted vehicle service recommendation according to a personalized scheme preset by the user.
In some embodiments, obtaining real-time user data and making vehicle service recommendations based on the real-time user data includes:
and monitoring vehicle environment data according to the real-time data of the user.
And combining the real-time data of the user and the vehicle environment data to sort the preset vehicle services to obtain sorting information.
And sequentially recommending the vehicle service to the preset vehicle service according to the sequencing information.
The real-time updating method provided by the application can also be used for recommending the vehicle services for a series of vehicle services according to a certain recommendation sequence. Specifically, the method can monitor vehicle environment data around the vehicle according to the user real-time data, sort the priority of each service in the preset vehicle service by combining the user real-time data and the vehicle environment data, and generate sorting information according to the sequence of the priority from high to low. By the method, each service in the preset vehicle service can be effectively sequenced according to the personalized requirements of the user and the current implementation condition of the vehicle, and the vehicle service recommendation efficiency is improved.
In some specific embodiments, the user real-time data includes historical information of the user in previous use processes, including execution services adopted in the past, context information, and the like. The vehicle environment data includes the current running state of the vehicle, the working condition of the vehicle itself, and history information of the vehicle recommendation service, etc.
In some embodiments, the user feedback data is derived from feedback recommended by the vehicle service, including:
and acquiring the first user feedback within the preset acquisition time length. The first user feedback comprises recommended services accepted by the user within a preset acquisition duration.
And acquiring second user feedback after the preset acquisition time length. And the second user feedback comprises the recommended service which is not responded or refused by the user within the preset acquisition time.
And continuously acquiring the third user feedback initiated by the user. Wherein the third user feedback comprises a service actively initiated by the user.
User feedback data is generated from the first user feedback, the second user feedback, and the third user feedback.
According to the method and the device, the behaviors of the user can be classified according to the preset acquisition duration of the user, so that the judgment of the effectiveness of the vehicle recommendation service by the user is judged, and the judgment is used as the user feedback data to optimize the initial recommendation model. Specifically, the method comprises the steps of obtaining first user feedback used for representing the judgment of the effectiveness and the accuracy of a user on recommended services within a preset obtaining time, collecting recommended services rejected by the user within the preset obtaining time after the preset obtaining time is exceeded, and obtaining second user feedback.
Specifically, the first user feedback and the second user feedback further include context information at corresponding moments, so that the method can more efficiently perform model updating on the initial recommendation model according to the user feedback data.
In some embodiments, the real-time updating method further comprises:
and if the user feedback data does not contain the second user feedback and the third user feedback, taking the initial recommendation model at the current moment as the service recommendation model.
The method can avoid updating of the initial recommendation model when the user completely receives the recommendation service provided by the initial recommendation model, and continuously recommend the vehicle service to the user according to the initial recommendation model. Specifically, when the user feedback data is analyzed and the second user feedback and the third user feedback are not obtained, it can be judged that the current initial recommendation model can meet the vehicle service recommendation requirements of the user. And the initial recommendation model is taken as a service recommendation model to continuously recommend the vehicle service, thereby avoiding unnecessary calculation load of the method on the system.
In some embodiments, the real-time update method further comprises:
and generating an initial Q network according to a preset network structure.
And obtaining an initial data set, and performing model training on the initial Q network according to the initial data set to obtain an initial recommendation model.
The method can also acquire the personalized data of the user according to the initial data set and sequentially provide the personalized dataAn initial recommendation model. Specifically, referring to fig. 2, the method generates an initial Q network by using the structure of fig. 2, and acquires training data in an initial data set to perform data training on the initial Q network, so as to obtain an initial training model. Specifically, the data stored in the initial dataset may be s t 、a t 、r t 、s t+1 A quadruple of four parameters. Wherein s is t Representing the current state ， s t+1 Represents the next state, a t Representing the currently executing service, r t Is in a state s t Lower execution service a t The latter value of return. According to the method, a plurality of quaternions in an initial data set are randomly adopted to train the initial Q network so as to obtain an initial recommendation model.
In some embodiments, the real-time update method further comprises:
and updating the initial data set according to the user feedback data to obtain an updated data set.
And according to the preset initialization duration, performing model training on the initial Q network through updating the data set so as to update the initial recommendation model.
According to the method and the device, the initial data set can be updated according to the user feedback data, so that the fitting degree of the initial data set and the user personalized requirements is improved, and the recommendation effectiveness of the initial recommendation model is further improved. In particular, the method may use the user feedback data as training parameters within the initial data set and generate an updated data set comprising the user feedback data. Further, after the preset initialization duration, model training is carried out on the initial Q network through updating the data set, updated model parameters are obtained and serve as the initial parameters to generate a new initial recommendation model.
In some real-time approaches, updating the initial data set according to the user feedback data to obtain an updated data set, including:
and calling the real-time user data at the next moment according to the user feedback data.
First scoring data is generated according to feedback of the user on the vehicle service recommendation.
And updating the initial data set according to the first grading data, the vehicle service recommendation, the user real-time data at the current moment and the user real-time data at the next moment to obtain an updated data set.
The method specifically comprises the steps of taking the user real-time data at the current moment, the user real-time data at the next moment, vehicle service recommendation and first scoring data generated according to feedback of the vehicle service recommendation as parameters of an updated initial data set, updating data in the initial data set according to the parameters to obtain an updated data set, and improving recommendation accuracy of an initial recommendation model through the updated data set. Specifically, the first score data may be generated according to the following rule. The method further adopts t As a feedback score to vehicle service recommendations. Specifically, when the user performs the recommended service, r t And =1. When the recommended service is ignored, r t And =0. When the recommended service is rejected, r t And (4) = -1. When a user actively uses a certain service in a certain scene at a certain time, r t =2, in order to calibrate the specific feedback status of the user therewith.
In some embodiments, updating the initial recommendation model according to the user feedback data to obtain a service recommendation model includes:
and grading the user feedback data according to a preset grading rule to obtain second grading data.
And updating the model parameters of the initial recommendation model according to the user feedback data and the second grading data to obtain a service recommendation model.
Similar to the scoring rules, the method can also update the initial recommendation model currently running according to the real-time feedback of the user during current vehicle running so as to provide a service recommendation model with higher fitting degree. Specifically, according to a preset grading rule, different weights are given to the user feedback data for accepting the recommendation service, rejecting the recommendation service, ignoring the recommendation service and initiatively initiating the service request, and the parameters of the initial recommendation model are trained according to the weights, so that the service recommendation model is generated by updating the parameters of the initial recommendation model.
Specifically, in addition to the second scoring data, the method may further use user personal data, such as user attributes (gender, age, and the like), vehicle interior and exterior conditions (temperature, humidity, and the like), vehicle operation attributes, and the like, as context states of the second scoring data, so that a service recommendation model with higher accuracy is recommended according to better fitting of the second scoring data and the context states.
Referring to fig. 3, the present application provides a real-time update apparatus for a vehicle service recommendation model, including:
and the online updating module is used for acquiring the real-time data of the user and recommending the vehicle service according to the real-time data of the user. And the method is also used for obtaining user feedback data according to the feedback recommended by the vehicle service. And the method is also used for reading the initial recommendation model at the current moment according to the user feedback data. And the method is also used for updating the initial recommendation model according to the user feedback data to obtain the service recommendation model.
The online updating module provided by the application can update the initial recommendation model according to the real-time feedback of the user in the vehicle running process so as to obtain the service recommendation model with higher recommendation accuracy. Specifically, by acquiring real-time data of the user, the device can recommend the vehicle service according to the preset recommendation service, acquire feedback of the user on the vehicle service recommendation, and generate user feedback data, so that model training is performed on the acquired initial recommendation model at the current moment according to the user feedback data to update the initial recommendation model, and the service recommendation model with higher recommendation accuracy is obtained.
Specifically, the device can further continue to obtain the real-time data of the user in real time after obtaining the service recommendation model so as to obtain the user feedback data, and further update the new initial recommendation model according to the user feedback data so as to update the new initial recommendation model again according to the current vehicle and the actual environment of the user so as to obtain the new service recommendation model. The new initial recommendation model may be a service recommendation model generated at a previous time, and the new service recommendation model may be an initial recommendation model at a next time.
Referring to fig. 3, in some embodiments, the apparatus further comprises:
and the offline updating module is used for generating an initial Q network according to a preset network structure. And the method is also used for obtaining an initial data set and carrying out model training on the initial Q network according to the initial data set to obtain an initial recommendation model. And the initial recommendation model is updated according to the user feedback data to obtain an updated data set, so that the initial recommendation model is updated according to the updated data set.
The device can preprocess the personalized information of the user through the offline updating module to obtain an initial recommendation model which can preliminarily represent the service preference of the user, and can update the initial data set according to user feedback data generated in the using process of vehicle service recommendation by the user so as to better provide the initial recommendation model with higher accuracy in advance. The real-time updating efficiency of the device to the initial recommendation model is greatly improved.
Referring to FIG. 4, FIG. 4 provides a real-time update process for the vehicle service recommendation model implemented cooperatively by the offline update module and the online update module. Wherein s is t As a vehicle service recommendation that the vehicle is performing in the current state, s t+1 As a vehicle service recommendation that the vehicle is performing in the next state, a t Representing the specific service currently being performed and w representing the parameters of the initial Q network. In particular, the current state s t As input to the initial Q network, the current state s is obtained t Lower optional service a-value list Q(s) t ,a t ,w)。
Further, denote the state by s, s t Representing the current state, s t+1 Represents the next state, s t ＝[s t 1 ,s t 2 ,s t 3 ,…,-s t 1 ,-s t 2 ,…,x t 1 ,x t 2 ,…,sex,age,temperature,humidity,time…]Wherein s is t i Indicating the ith recommended by the system to the user and the user receiving an action (service), -s t i Then an action (service), x, is indicated that the system recommends to the user but is rejected by the user t i Actions (services) not recommended on behalf of the system but actively used by the user) The latter signals in the list represent the context state at time t, such as the owner's attributes (gender, age, etc.), the conditions inside and outside the vehicle (temperature, humidity, etc.), and the attributes of the vehicle itself.
Further, the service a recommended by the network when Q is started t Accepted by the user, and the initial Q network does not make any update at this time. Service a recommended when initial Q network t Refused by the user or actively used by the user in the absence of a recommended service t According to user feedback r t And the next state s t+1 The network parameter w is updated.
Further,(s) t ,a t ,r t ,s t+1 ) Middle r t Is in a state s t Lower execution service a t Later reward, when the user performs the recommended service, r t And =1. When the recommended service is ignored, r t And =0. When the recommended service is rejected, r t And (5) keeping the value of-1. When a user actively uses a certain service in a certain scene at a certain time, r t ＝2。s t+1 Is represented in state s t Lower execution service a t The next state is obtained.
Further, when the system recommends a service to the user through the initial Q network, if the service is ignored, rejected, or the user actively performs some service, the parameter training is performed, and then the initial Q network is updated.
The initial data set may also be in the above state s t Just performing service a t Then store the resulting reward r t And the next state s t+1 A quadruple of (i),(s) t ,a t ,r t ,s t+1 )。
Referring to fig. 5, the present application provides a vehicle comprising: a memory, a processor and a program stored on the memory and executable on the processor, the processor executing the program to implement a method of real-time updating of a vehicle service recommendation model as in any one of the above.
By executing the program stored in the memory through the processor provided by the application, the vehicle provided by the application can realize the real-time updating method of the vehicle service recommendation model. The foregoing has been described in detail with respect to the technical effects that can be provided by implementing the above updating method, and no further description is provided herein.
The present application provides a computer-readable storage medium having stored thereon a computer program for execution by a processor for implementing a method of real-time updating of a vehicle service recommendation model according to any preceding claim.
The method for updating the vehicle service recommendation model in real time can be realized by executing the computer program stored in the computer readable storage medium provided by the application. The foregoing has been described in detail with respect to the technical effects that can be provided by implementing the above updating method, and no further description is provided herein.
In the description of the present specification, reference to the description of "one embodiment," "some embodiments," "an example," "a specific example," or "some examples" or the like means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or N embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the feature. In the description of the present application, "N" means at least two, e.g., two, three, etc., unless explicitly defined otherwise.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more N executable instructions for implementing steps of a custom logic function or process, and alternate implementations are included within the scope of the preferred embodiment of the present application in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of implementing the embodiments of the present application.
It should be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: discrete logic circuits having logic gates for implementing logic functions on data signals, application specific integrated circuits having suitable combinational logic gates, programmable gate arrays, field programmable gate arrays, and the like.
It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware that is related to instructions of a program, and the program may be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.
While embodiments of the present application have been shown and described above, it will be understood that the above embodiments are exemplary and should not be construed as limiting the present application and that changes, modifications, substitutions and alterations in the above embodiments may be made by those of ordinary skill in the art within the scope of the present application.
Claims (12)
1. A real-time updating method for a vehicle service recommendation model is characterized by comprising the following steps:
acquiring user real-time data, and recommending vehicle service according to the user real-time data;
obtaining user feedback data according to the feedback recommended by the vehicle service;
reading an initial recommendation model at the current moment according to the user feedback data;
and updating the initial recommendation model according to the user feedback data to obtain a service recommendation model.
2. The real-time updating method of claim 1, wherein the obtaining of the real-time user data and the recommendation of the vehicle service according to the real-time user data comprises:
monitoring vehicle environment data according to the user real-time data;
combining the user real-time data and the vehicle environment data to sort preset vehicle services to obtain sorting information;
and sequentially recommending the vehicle service to the preset vehicle service according to the sequencing information.
3. The real-time update method of claim 1, wherein the obtaining user feedback data based on the feedback recommended by the vehicle service comprises:
acquiring first user feedback within a preset acquisition time length; the first user feedback comprises recommended services accepted by the user within the preset acquisition duration;
after the preset acquisition duration, acquiring second user feedback; the second user feedback comprises recommended services which are not responded or refused by the user within the preset acquisition duration;
continuously acquiring third user feedback initiated by the user; wherein the third user feedback comprises a service actively initiated by the user;
generating the user feedback data from the first user feedback, the second user feedback, and the third user feedback.
4. The real-time update method of claim 3, further comprising:
and if the user feedback data does not contain the second user feedback and the third user feedback, taking the initial recommendation model at the current moment as a service recommendation model.
5. The real-time update method of claim 3, further comprising:
generating an initial Q network according to a preset network structure;
and obtaining an initial data set, and performing model training on the initial Q network according to the initial data set to obtain the initial recommendation model.
6. The real-time update method of claim 5, further comprising:
updating the initial data set according to the user feedback data to obtain an updated data set;
and performing model training on the initial Q network through the updating data set according to a preset initialization duration so as to update the initial recommendation model.
7. The real-time updating method of claim 6, wherein the updating the initial data set according to the user feedback data to obtain an updated data set comprises:
calling the real-time user data at the next moment according to the user feedback data;
generating first scoring data according to feedback of a user on the vehicle service recommendation;
and updating the initial data set according to the first grading data, the vehicle service recommendation, the user real-time data at the current moment and the user real-time data at the next moment to obtain an updated data set.
8. The real-time updating method of claim 1, wherein the updating the initial recommendation model according to the user feedback data to obtain a service recommendation model comprises:
grading the user feedback data according to a preset grading rule to obtain second grading data;
and updating the model parameters of the initial recommendation model according to the user feedback data and the second grading data to obtain a service recommendation model.
9. An apparatus for real-time update of a vehicle service recommendation model, comprising:
the online updating module is used for acquiring real-time user data and recommending vehicle service according to the real-time user data; the vehicle service recommendation system is also used for obtaining user feedback data according to the feedback recommended by the vehicle service; the system is also used for reading the initial recommendation model at the current moment according to the user feedback data; and the initial recommendation model is updated according to the user feedback data to obtain a service recommendation model.
10. The real-time update apparatus of claim 9, wherein the apparatus further comprises:
the offline updating module is used for generating an initial Q network according to a preset network structure; the system is also used for obtaining an initial data set and carrying out model training on the initial Q network according to the initial data set to obtain the initial recommendation model; and the initial recommendation model is updated according to the user feedback data to obtain an updated data set, so that the initial recommendation model is updated according to the updated data set.
11. A vehicle, characterized by comprising: memory, processor and program stored on the memory and executable on the processor, the processor executing the program to implement a method of real-time updating of a vehicle service recommendation model according to any of claims 1-8.
12. A computer-readable storage medium, having stored thereon a computer program, the program being executable by a processor for implementing a method for real-time update of a vehicle service recommendation model according to any one of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211018341.2A CN115470403A (en) | 2022-08-24 | 2022-08-24 | Real-time updating method and device of vehicle service recommendation model, vehicle and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211018341.2A CN115470403A (en) | 2022-08-24 | 2022-08-24 | Real-time updating method and device of vehicle service recommendation model, vehicle and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115470403A true CN115470403A (en) | 2022-12-13 |
Family
ID=84366744
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211018341.2A Pending CN115470403A (en) | 2022-08-24 | 2022-08-24 | Real-time updating method and device of vehicle service recommendation model, vehicle and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115470403A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116016658A (en) * | 2023-01-05 | 2023-04-25 | 中国第一汽车股份有限公司 | Recommendation method of vehicle service and vehicle |
-
2022
- 2022-08-24 CN CN202211018341.2A patent/CN115470403A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116016658A (en) * | 2023-01-05 | 2023-04-25 | 中国第一汽车股份有限公司 | Recommendation method of vehicle service and vehicle |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10402039B2 (en) | Adaptive user interface using machine learning model | |
| EP3711000B1 (en) | Regularized neural network architecture search | |
| US11521056B2 (en) | System and methods for intrinsic reward reinforcement learning | |
| CN111813921B (en) | Topic recommendation method, electronic device and computer-readable storage medium | |
| CN109062919B (en) | Content recommendation method and device based on deep reinforcement learning | |
| EP3583553A1 (en) | Neural architecture search for convolutional neural networks | |
| TW201923675A (en) | Information recommendation method, device and apparatus | |
| EP3648011A1 (en) | Hierarchical clustered reinforcement machine learning | |
| CN110826725A (en) | Intelligent agent reinforcement learning method, device and system based on cognition, computer equipment and storage medium | |
| CN115470403A (en) | Real-time updating method and device of vehicle service recommendation model, vehicle and medium | |
| WO2021202528A1 (en) | Multi-modal neural network architecture search | |
| WO2022005626A1 (en) | Partially-observed sequential variational auto encoder | |
| Fanjiang et al. | Time series QoS forecasting for Web services using multi-predictor-based genetic programming | |
| JP7348230B2 (en) | Generation device, generation method, and generation program | |
| Zhang et al. | Home health care routing problem via off-line learning and neural network | |
| CN110717537B (en) | Method and device for training user classification model and executing user classification prediction | |
| CN110070104B (en) | User recommendation method, device and server | |
| CN111046156A (en) | Method and device for determining reward data and server | |
| CN113449176A (en) | Recommendation method and device based on knowledge graph | |
| JP7466702B2 (en) | Interpretable Imitation Learning by Discovering Prototype Options | |
| CN114723074A (en) | Active learning client selection method and device under clustering federal learning framework | |
| CN112760908A (en) | Laundry system and control method | |
| CN111061875B (en) | Super parameter determination method, device, computer equipment and storage medium | |
| CN112070205A (en) | Multi-loss model obtaining method and device | |
| CN111401594A (en) | Display information combination determination method and device, storage medium and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |