Vehicle content recommendation method and system, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a vehicle content recommendation method and system, electronic equipment and a storage medium, and belongs to the technical field of computers. According to the method, user feature data and user behavior data are acquired, vehicle type preference prediction is performed according to the user feature data and the user behavior data by using a deep learning technology, a vehicle type preferred by a user is obtained, and a recommended test driving site is determined according to user position information and the vehicle type. And then obtaining a vehicle type set corresponding to the vehicle type and performance index preference weight information of the user, calculating a matching degree of each vehicle type in the vehicle type set according to the performance index preference weight information, and determining recommended vehicle content according to the vehicle type, the recommended test driving place and the matching degree of each vehicle type in the vehicle type set. According to the invention, comprehensive vehicle content recommendation conforming to user preferences can be provided for the user, and the user experience is improved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a vehicle content recommendation method, system, electronic device and storage medium. Background Art

[0002] In the modern automobile industry, consumers are faced with a large amount of information and choices when choosing a car. With the rapid development of information technology, especially the popularity of the Internet and mobile applications, consumers have a variety of ways to obtain automobile information, which enables them to access a large amount of vehicle information, brands and services when choosing a car. However, as the range of choices expands, the problem of information overload faced by consumers when making purchasing decisions has become more and more obvious.

[0003] In the related art, vehicle recommendations can be made based on the user's historical browsing information, but the recommended content is too single, and generally only certain specific brands, series and models of vehicles are recommended, resulting in too little selectivity. The recommended vehicles are difficult to meet user needs, and it is impossible to provide users with reliable and comprehensive vehicle content recommendations, which affects the user experience. Summary of the invention

[0004] The main purpose of the embodiments of the present application is to propose a vehicle content recommendation method, system, electronic device and storage medium, aiming to provide users with comprehensive vehicle content recommendations that meet user preferences and improve user experience.

[0005] To achieve the above purpose, one aspect of an embodiment of the present application proposes a vehicle content recommendation method, comprising the following steps:

[0006] Obtain user feature data and user behavior data;

[0007] Using deep learning technology, predicting vehicle type preference based on the user feature data and the user behavior data, and obtaining the vehicle type preferred by the user;

[0008] Obtaining user location information, and determining a recommended test drive location based on the user location information and the vehicle type;

[0009] Acquire a vehicle model set corresponding to the vehicle type and the user's performance indicator preference weight information, and calculate the matching degree of each vehicle model in the vehicle model set according to the performance indicator preference weight information;

[0010] The recommended vehicle content is determined according to the vehicle type, the recommended test drive location, and the matching degree of each vehicle model in the vehicle model set.

[0011] In some embodiments, the use of deep learning technology to predict vehicle type preferences based on the user feature data and the user behavior data to obtain the vehicle type preferred by the user includes the following steps:

[0012] Preprocessing the user feature data and the user behavior data to obtain data to be analyzed;

[0013] Extracting features of the data to be analyzed to obtain basic user features, browsing features, operation features, and comment content features;

[0014] The format-converted user basic features, the browsing features, the operation features, and the comment content features are input into a user preference prediction model built based on deep learning technology to obtain the vehicle type preferred by the user.

[0015] In some embodiments, determining a recommended test drive location based on the user location information and the vehicle type includes the following steps:

[0016] Obtaining a preset first test drive information table, wherein the first test drive information table includes vehicle inventory information of each test drive location;

[0017] Filtering the test drive locations in the first test drive information table according to the user location information to obtain first candidate locations, and forming a second test drive information table according to the vehicle inventory information of each of the first candidate locations;

[0018] Retrieving the vehicle inventory information in the second test drive information table according to the vehicle type to obtain a second candidate location where a vehicle corresponding to the vehicle type is stored;

[0019] The second candidate location is determined as a recommended test drive location.

[0020] In some embodiments, the step of filtering the test drive locations in the first test drive information table according to the user location information to obtain a first candidate location includes the following steps:

[0021] Calculating the distance between the user location and each test drive location in the first test drive information table according to the user location information and the location information of the test drive location;

[0022] A test drive location whose distance is less than a first threshold is selected as a first candidate location.

[0023] In some embodiments, the performance indicator preference weight information is obtained by the following steps:

[0024] Get the user's historical browsing data for a preset time period before the current time;

[0025] Extracting target content from the historical browsing data, wherein the target content includes semantic keywords related to performance indicators, browsed vehicle models, and dwell time of browsed vehicle models;

[0026] Acquire performance characteristics of each browsed vehicle model, wherein the performance characteristics include at least one performance index;

[0027] According to the stay time of the browsed vehicle model, the total time of each performance indicator in the performance characteristics of all the browsed vehicle models is counted, and the first weight of each performance indicator is determined according to the total time of each performance indicator;

[0028] Counting the number of occurrences of the semantic keyword of each performance indicator in the target content, and determining a second weight of the performance indicator according to the number of occurrences;

[0029] The user's performance indicator preference weight information is determined according to the first weight and the second weight of each performance indicator.

[0030] In some embodiments, the calculating the matching degree of each vehicle model in the vehicle model set according to the performance indicator preference weight information comprises the following steps:

[0031] Acquire performance characteristics of each vehicle model in the vehicle model set, wherein the performance characteristics include at least one performance indicator;

[0032] determining a weight of each performance indicator in the performance characteristics of the vehicle model according to the performance indicator preference weight information;

[0033] The matching degree of the vehicle model is determined according to the weights of all performance indicators in the performance characteristics of the vehicle model.

[0034] In some embodiments, the vehicle content recommendation method further includes the following steps:

[0035] In response to a user's feedback operation on the vehicle content, updating the user behavior data;

[0036] Generate new vehicle content based on the updated user behavior data.

[0037] To achieve the above purpose, another aspect of the embodiment of the present application provides a vehicle content recommendation system, including:

[0038] The first module is used to obtain user feature data and user behavior data;

[0039] The second module is used to use deep learning technology to predict vehicle type preference based on the user feature data and the user behavior data to obtain the vehicle type preferred by the user;

[0040] The third module is used to obtain user location information and determine a recommended test drive location based on the user location information and the vehicle type;

[0041] The fourth module is used to obtain the vehicle model set corresponding to the vehicle type and the user's performance indicator preference weight information, and calculate the matching degree of each vehicle model in the vehicle model set according to the performance indicator preference weight information;

[0042] The fifth module is used to determine the recommended vehicle content according to the vehicle type, the recommended test drive location and the matching degree of each vehicle model in the vehicle model set.

[0043] To achieve the above-mentioned purpose, another aspect of an embodiment of the present application proposes an electronic device, which includes a memory, a processor, a program stored in the memory and executable on the processor, and a data bus for realizing connection and communication between the processor and the memory, and the program implements the method described in the above embodiment when executed by the processor.

[0044] To achieve the above-mentioned purpose, another aspect of an embodiment of the present application proposes a storage medium, which is a computer-readable storage medium used for computer-readable storage, and the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the method described in the above embodiment.

[0045] The vehicle content recommendation method, system, electronic device and storage medium proposed in the present application obtain user feature data and user behavior data, use deep learning technology, predict vehicle model preferences based on user feature data and user behavior data, obtain the vehicle type preferred by the user, and determine the recommended test drive location based on the user location information and vehicle type, then obtain the vehicle model set corresponding to the vehicle type and the user's performance index preference weight information, and calculate the matching degree of each vehicle model in the vehicle model set based on the performance index preference weight information, and determine the recommended vehicle content based on the vehicle type, the recommended test drive location and the matching degree of each vehicle model in the vehicle model set. The present application can provide users with comprehensive vehicle content recommendations that meet user preferences and improve user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is a flow chart of a vehicle content recommendation method provided by an embodiment of the present application;

[0047] Figure 2 It is a schematic diagram of the deep learning model structure design process provided by the embodiment of the present application;

[0048] Figure 3 It is a schematic diagram of the deep learning model update optimization process provided by an embodiment of the present application;

[0049] Figure 4 It is a schematic diagram of the hardware structure of the electronic device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0051] It should be noted that, although the functional modules are divided in the system and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the system or the order in the flowchart. The terms "first", "second", etc. in the specification, claims and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0052] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0053] Based on this, the embodiments of the present application provide a vehicle content recommendation method, system, electronic device and storage medium, which aim to provide users with comprehensive vehicle content recommendations that meet user preferences and improve user experience.

[0054] The vehicle content recommendation method, system, electronic device and storage medium provided in the embodiments of the present application are specifically described through the following embodiments. First, the vehicle content recommendation method in the embodiments of the present application is described.

[0055] The vehicle content recommendation method provided in the embodiment of the present application relates to the field of computer technology. The vehicle content recommendation method provided in the embodiment of the present application can be applied to a terminal, can be applied to a server side, or can be software running in a terminal or a server side. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc.; the server side can be configured as an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application that implements the vehicle content recommendation method, etc., but is not limited to the above forms.

[0056] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0057] It should be noted that in each specific implementation of the present application, when it comes to the need to perform relevant processing based on data related to user identity or characteristics such as user information, user behavior data, user historical data, and user location information, the user's permission or consent will be obtained first, and the collection, use, and processing of these data will comply with relevant laws, regulations, and standards. In addition, when the embodiment of the present application needs to obtain the user's sensitive personal information, the user's separate permission or consent will be obtained through a pop-up window or by jumping to a confirmation page. After clearly obtaining the user's separate permission or consent, the necessary user-related data for the normal operation of the embodiment of the present application will be obtained.

[0058] Figure 1 is an optional flowchart of the vehicle content recommendation method provided in an embodiment of the present application. Figure 1 The method may include but is not limited to steps S101 to S105.

[0059] Step S101, obtaining user feature data and user behavior data;

[0060] Step S102, using deep learning technology to predict vehicle type preference based on user feature data and user behavior data to obtain the vehicle type preferred by the user;

[0061] Step S103, obtaining user location information, and determining a recommended test drive location based on the user location information and the vehicle type;

[0062] Step S104, obtaining a vehicle model set corresponding to the vehicle type and the user's performance index preference weight information, and calculating the matching degree of each vehicle model in the vehicle model set according to the performance index preference weight information;

[0063] Step S105 , determining the recommended vehicle content according to the vehicle type, the recommended test drive location, and the matching degree of each vehicle model in the vehicle model set.

[0064] In step S101 of some embodiments, user feature data refers to basic information of the user, such as age, residential area, occupation, etc. User behavior data refers to data related to user interaction operations on the front end, including but not limited to browsing history, operation behavior, comment data, and purchase behavior, etc. Accurate prediction of user preference vehicles can be achieved by obtaining rich user data (including user feature data and user behavior data).

[0065] In step S102 of some embodiments, a user preference prediction model is constructed using deep learning technology. The user preference prediction model can reflect the mapping relationship between user data and vehicle type labels. The model can use various neural network structures of deep learning (such as DNN, CNN, RNN). After obtaining the mapping relationship between user data and vehicle type labels, the mapping relationship is used to predict the vehicle type preference of the user feature data and user behavior data of the current user to obtain the vehicle type preferred by the user. Vehicle types include but are not limited to sedans, SUVs (sports utility vehicles), MPVs (multi-purpose commercial vehicles), sports cars, minivans (vans), off-road vehicles and buses, etc. It can be understood that the user prediction model trained based on deep learning is a classification model. Using this model to predict user data can obtain the prediction results of the probability values ​​of multiple vehicle types, and based on the prediction results, a list of vehicle types with the top probability values ​​can be generated. This embodiment can provide effective recommendations in the case of sparse data compared to the collaborative filtering recommendation algorithm through the effectiveness and powerful feature extraction capabilities of the deep learning algorithm, and can also make effective recommendations for new users and new models.

[0066] In some embodiments, step S102 may include but is not limited to steps S201 to S203:

[0067] Step S201, pre-processing the user feature data and the user behavior data to obtain data to be analyzed;

[0068] Step S202, extracting features from the data to be analyzed to obtain user basic features, browsing features, operation features, and comment content features;

[0069] Step S203, inputting the format-converted user basic features, browsing features, operation features, and comment content features into a user preference prediction model built based on deep learning technology to obtain the vehicle type preferred by the user.

[0070] In this embodiment, the collected user feature data and user behavior data are cleaned, standardized and other pre-processing operations are performed to obtain the data to be analyzed. The natural language data technology is used to extract features of the data to be analyzed to obtain user basic features, browsing features, operation features and comment content features, and then the extracted features are formatted to conform to the model input format, and the format-converted user basic features, browsing features, operation features and comment content features are input into the user preference prediction model built based on deep learning technology to obtain the vehicle type preferred by the user.

[0071] Exemplarily, the input data processing process of the deep learning model (i.e., the user preference prediction model) is as follows:

[0072] S11, class initialization:

[0073] A class named RecommendationSystem is defined. In its initialization method, an empty user data list is created. User_data is used to store user behavior data. At the same time, a recommendation model object RecommendationModel is initialized for subsequent recommendation processing.

[0074] S12. User data collection:

[0075] Use the collect_user_data method to collect user behavior data. This method receives a parameter user_action, which represents the user's specific operation or behavior, and adds it to the user_data list.

[0076] S13. Data preprocessing:

[0077] The preprocess_data method is used to preprocess the collected user data. Assume that the user data is stored in the form of DataFrame. The main steps of this method include:

[0078] Data cleaning: Remove duplicate entries to ensure data uniqueness and accuracy.

[0079] Feature extraction: Call the feature_extraction method to extract and process features from the cleaned data. The main function of the feature_extraction method is to extract features from the cleaned data and standardize them. In this method, the features are standardized by using StandardScaler to ensure that the values ​​of different features are in the same range, thereby improving the effect of model learning or model prediction. The objects of feature extraction include the user's age, budget, previously used car models, and browsing content.

[0080] For example, please refer to Figure 2 , the adaptive deep learning model structure used in the embodiment of the present application is described as follows:

[0081] Input Layer:

[0082] The input features include basic user features (such as age, gender, budget, and geographic location), browsing features (such as browsing history), operation features (such as number of clicks), and comment content features.

[0083] Hidden Layer:

[0084] The hidden layer abstracts and transforms the input features layer by layer through a multi-layer neural network. Each layer can capture feature information at different levels, and the complexity and expression ability of the model increase with the increase of the number of layers.

[0085] Output layer:

[0086] The output layer generates a recommendation probability distribution for each user and recommends several vehicle types.

[0087] The specific recommendation implementation process of the deep learning model is as follows:

[0088] S21. Class definition:

[0089] Define a class called RecommendationModel that inherits from TensorFlow's tf.keras.Model. This class builds a deep learning model for recommendation.

[0090] S22, initialization method:

[0091] In the initialization method __init__, call the super() constructor of the parent class to correctly initialize the base class. Then define three fully connected layers (Dense layers):

[0092] The first layer (dense1): contains 128 neurons and uses ReLU (rectified linear unit) as the activation function. This layer is responsible for extracting high-order features of the input features.

[0093] The second layer (dense2): contains 64 neurons and also uses the ReLU activation function to further process and refine the features from the first layer.

[0094] Output layer (output_layer): contains 10 neurons and uses the softmax activation function. The output of this layer represents the model's recommendation probability for 10 possible car models, which is suitable for multi-class classification problems.

[0095] S23, forward propagation (call method):

[0096] The call method defines the forward propagation process of the model. This method receives the input data inputs and passes through each layer in order. The input data first passes through the first layer dense1, and the output is passed to the second layer dense2. Finally, the output result of the second layer is sent to the output layer output_layer to generate the final recommendation result, which indicates the probability of each vehicle type recommended to the user.

[0097] Exemplarily, the data processing flow of the embodiment of the present application is described. The data processing flow can be used to process input data of the model training process, the model prediction process and the model update process. The processing flow mainly includes:

[0098] Data cleaning and preprocessing: remove duplication and normalize data to improve data quality and consistency.

[0099] Feature extraction and conversion: Extract user features and behavior features from the cleaned data and convert them into a format that can be used for model input.

[0100] Processed data stream: Generate cleaned and processed data stream for subsequent deep learning model training and prediction.

[0101] The specific processing flow is described as follows:

[0102] S31, data processing function:

[0103] A function named data_processing is defined, which is responsible for the entire data processing process, including data cleaning and feature extraction. It receives user data user_data as input and finally returns the extracted features.

[0104] S32. Data cleaning:

[0105] Inside the data_processing function, the clean_data function is first called to clean the input data. The cleaning operation mainly involves removing duplicate entries to ensure the uniqueness and accuracy of the data. The cleaned data is stored in the cleaned_data variable.

[0106] S33, feature extraction:

[0107] Call the extract_features function to extract features from the cleaned data. The main task of this function is to extract relevant features from the cleaned data and perform standardization. The objects of feature extraction include the user's age, budget, and previously used car model. In the extract_features function, StandardScaler is used to standardize the selected features so that the values ​​of different features are in the same range, which helps to improve the learning effect of the subsequent model.

[0108] S34, return characteristics:

[0109] The data_processing function returns the extracted and standardized feature data for subsequent model training and recommendation.

[0110] In step S103 of some embodiments, after determining the vehicle type preferred by the user, the test drive location of the vehicle type can be determined near the user in combination with the user's location information. The user can directly obtain the test drive location of the preferred vehicle type, reducing the user's self-location search operations and improving the user experience.

[0111] In some embodiments, step S103 may include but is not limited to steps S301 to S304:

[0112] Step S301, obtaining a preset first test drive information table, wherein the first test drive information table includes vehicle inventory information of each test drive location;

[0113] Step S302, screening the test drive locations in the first test drive information table according to the user location information to obtain first candidate locations, and forming a second test drive information table according to the vehicle inventory information of each first candidate location;

[0114] Step S303, searching the vehicle inventory information in the second test drive information table according to the vehicle type to obtain a second candidate location where a vehicle corresponding to the vehicle type is stored;

[0115] Step S304: determine the second candidate location as the recommended test drive location.

[0116] In this embodiment, the first test drive information table includes the vehicle inventory information of each test drive location (dealer), and the vehicle inventory information includes the vehicle type, vehicle model and other information of the test drive location. The test drive locations in the first test drive information table are screened according to the user location information to obtain the test drive locations near the user (i.e., the first candidate locations), and the second test drive information table is formed according to the vehicle inventory information of each first candidate location, and then the vehicle inventory information in the second test drive information table is retrieved according to the vehicle type to obtain the second candidate location where the vehicle corresponding to the vehicle type is stored, and the second candidate location is the recommended test drive location. Furthermore, after obtaining multiple second candidate locations, the multiple second candidate locations can be sorted from high to low according to the user ratings of the dealers so as to recommend suitable test drive locations to the user.

[0117] In some embodiments, step S302 may include but is not limited to steps S401 to S402:

[0118] Step S401, calculating the distance between the user location and each test drive location in the first test drive information table according to the user location information and the location information of the test drive location;

[0119] Step S402: Select a test drive location whose distance is less than a first threshold as a first candidate location.

[0120] In this embodiment, the straight-line distance between the user location and each test drive location in the first test drive information table can be calculated based on the user location information and the location information of the test drive location, and the test drive location with a distance less than a first threshold (for example, 10 kilometers) is selected as the first candidate location, thereby screening out the test drive locations near the user.

[0121] In some embodiments, the user location information may refer to the user's real-time location or the user's residential address. The user's residential address may be obtained through the user's registration information, and the user's real-time location may be obtained through GPS, IP address, etc. with the user's consent.

[0122] In one example, when a user logs in or browses a web page, the system automatically obtains the user's geographic location and passes the information to the recommendation engine, which uses the geographic coordinate system to calculate the distance between the user's geographic location and nearby dealers. For example, the "Haversine formula" can be used to calculate the distance between two points on a sphere. The recommendation engine filters out the dealers closest to the user based on the type of vehicle preferred by the user and its distance from the user's geographic location. For example, if the user shows interest in "Brand A" SUV, the system will query the "Brand A" SUV dealer closest to the user. Further, the front end can also recommend to the user announcement information about the user's preferred vehicle type from nearby dealers, that is, the announcement information is added to the vehicle content and displayed on the front end. The announcement information can be set by the dealer manager logging into the recommendation system, for example, "The dealer near you is holding a special promotion for 'Brand A' SUV, only 5 kilometers away from you!". Further, on the recommended vehicle content, the distance information can be marked at the recommended test drive location to improve the user experience. The test drive location can be displayed on a map, that is, the user's current location and the location of the nearby dealer are marked on the map, so that the user can intuitively see the distance and the convenience of purchase.

[0123] In one example, when the user preference is not obvious, that is, the maximum probability value of the vehicle type output by the user preference prediction model does not exceed the second threshold, it is considered that the user preference is not obvious. At this time, dealers and models can be recommended directly based on the user's location. For example, test drive locations of several dealers within a certain range from the user are selected, and the vehicle type with the largest inventory is selected based on the dealer's vehicle inventory information and recommended to the user.

[0124] In step S104 of some embodiments, for a certain brand, a series of vehicle models will be designed for a certain vehicle type, which is called a car series. That is, a car series refers to a collection of models with similar characteristics or designs (i.e., the same vehicle type) under the same car brand. Different vehicle models under a car series have different performance indicators such as configuration, price, displacement, interior, safety performance and fuel consumption. Based on this, this embodiment, on the basis of determining the vehicle type preferred by the user, further combines the user's performance indicator preference weight information to calculate the matching degree of different vehicle models under the vehicle type car series, thereby providing users with accurate vehicle model recommendations.

[0125] In some embodiments, the performance indicator preference weight information in step S104 may be obtained by the following steps:

[0126] Step S501, obtaining historical browsing data of the user in a preset time period before the current time;

[0127] Step S502, extracting target content from the historical browsing data, wherein the target content includes semantic keywords related to performance indicators, browsed vehicle models, and the dwell time of the browsed vehicle models;

[0128] Step S503, obtaining performance characteristics of each browsed vehicle model, wherein the performance characteristics include at least one performance index;

[0129] Step S504, according to the stay time of the browsed vehicle model, the total time of each performance indicator in the performance characteristics of all browsed vehicle models is counted, and the first weight of each performance indicator is determined according to the total time of each performance indicator;

[0130] Step S505, counting the number of occurrences of the semantic keyword of each performance indicator in the target content, and determining a second weight of the performance indicator according to the number of occurrences;

[0131] Step S506: Determine the user's performance indicator preference weight information according to the first weight and the second weight of each performance indicator.

[0132] In this embodiment, the user's historical browsing data of a preset time period before the current time, for example, the historical browsing data of the last 10 days, can be used to analyze the user's performance indicator preference weight information based on the historical browsing data. The user's performance indicator preference weight information refers to the user's preference for performance indicators such as space size, fuel consumption, and safety.

[0133] In this embodiment, after obtaining the user's historical browsing data, the target content in the historical browsing data can be extracted, and the target content includes semantic keywords about performance indicators, browsed models, and the dwell time of the browsed models. For example, the user browsed a certain vehicle model of SUV many times in the past week, with an average dwell time of 8 minutes, and browsed a total of 5 related pages (including performance evaluation, user comments, price information, etc.). After browsing the vehicle model, the user clicked the "user comments" link and viewed 3 comments. In the comments, the user frequently liked the "large space" and "low fuel consumption" of the model. The user made a positive comment on the SUV, mentioning that "it is very suitable for family use and the travel space is satisfactory", and raised questions about "safety". Relevant content is extracted from these data, including semantic keywords about performance indicators such as space, fuel consumption and safety, browsed models (a certain vehicle model) and dwell time.

[0134] In this embodiment, after determining the model that the user browsed in the past, the performance characteristics of the model can be matched in the system. For example, safety, space and other aspects are the advantages of the browsed model. It can be considered that the reason why the user browses the model is the performance indicators of safety and space. According to the residence time of each browsed model, the total time of each performance indicator in the performance characteristics of all browsed models is counted. For example, the browsing time of model 1 is 6 minutes, and its performance characteristics include safety and space. The browsing time of model 2 is 8 minutes, and its performance characteristics include fuel consumption and safety. By counting the browsing time of model 1 and model 2, it can be obtained that the total time of the performance indicator of safety is 14 minutes. Further, the first weight of each performance indicator is determined according to the total time of each performance indicator. For example, the total time of each performance indicator is divided by the sum of the residence time of all browsed models to obtain the first weight of the performance indicator.

[0135] In this embodiment, the number of occurrences of the semantic keyword of each performance indicator in the target content is counted, and the second weight of the performance indicator is determined based on the number of occurrences. For example, the number of occurrences of the semantic keyword about fuel consumption in other people's comments liked by the user, comments posted by the user, etc. is 5 times. The number of occurrences of the semantic keyword is divided by the total number of occurrences of all semantic keywords to obtain the second weight of the performance indicator corresponding to the semantic keyword.

[0136] In this embodiment, the first weight and the second weight of each performance indicator are combined to determine the user's performance indicator preference weight information. For example, the first weight and the second weight of each performance indicator are weighted to obtain the weight of each performance indicator, and the weights of all performance indicators form the performance indicator preference weight information.

[0137] In some embodiments, the step of calculating the matching degree of each vehicle model in the vehicle model set according to the performance indicator preference weight information in step S104 includes but is not limited to steps S601 to S603:

[0138] Step S601, obtaining performance characteristics of each vehicle model in the vehicle model set, wherein the performance characteristics include at least one performance index;

[0139] Step S602, determining the weight of each performance indicator in the performance characteristics of the vehicle model according to the performance indicator preference weight information;

[0140] Step S603, determining the matching degree of the vehicle model according to the weights of all performance indicators in the performance characteristics of the vehicle model.

[0141] In this embodiment, the vehicle model set corresponding to the vehicle type preferred by the user is obtained, and the performance characteristics of each vehicle model in the vehicle model set are obtained. Then, based on the weight of each performance indicator in the performance indicator preference weight information, the weight of each performance indicator in the performance characteristics of the current vehicle model can be directly determined. For example, the performance characteristics of vehicle model 1 include safety and fuel consumption. Based on the performance indicator preference weight information, the weights of safety and fuel consumption are 0.1 and 0.2 respectively. The weights of each performance indicator in the performance characteristics of the current vehicle model are directly added together, and the matching degree of the current vehicle model is 0.3. The matching degree of a vehicle model represents the user's preference for the vehicle model.

[0142] For example, the specific process recommended for the vehicle model is as follows:

[0143] Vehicle model classification: All vehicle models in the automobile market are classified based on vehicle type, ensuring that each vehicle model can be attributed to the corresponding vehicle series, thus forming a vehicle model set of the vehicle series. For example, brand A's SUV may have multiple vehicle models, such as "Brand A Vehicle Model 1", "Brand A Vehicle Model 2", etc.

[0144] Index characteristic analysis of vehicle models under a car series: Index characteristic analysis is conducted on each vehicle type under each car series, including power system, interior configuration, space size, safety performance, etc., so as to form the index characteristics of each vehicle model.

[0145] User performance indicator preference weight analysis: By analyzing the user's historical browsing behavior, the performance indicators that the user is concerned about are analyzed and the appropriate weights are calculated. The performance indicator weight can reflect the user's attention to the performance indicator. For example, the user prefers high-configuration and large-space performance indicators, and the weight of this performance indicator is higher.

[0146] Recommendation of each vehicle model in the vehicle model set corresponding to the vehicle type: Based on the analyzed performance index preference weight information, the matching degree between each vehicle model and the user's preference can be calculated. Users prefer high-configuration and large-space performance indicators, and the system will give priority to recommending vehicle models with more matching configurations in the car series.

[0147] In step S105 of some embodiments, the recommended vehicle content is determined according to the vehicle type, the recommended test drive location, and the matching degree of each vehicle model in the vehicle model set. For example, the vehicle content can be displayed in a list format. The first column can display at least one vehicle type preferred by the user, and the second column can display the recommended test drive location corresponding to the vehicle type. If there are multiple recommended test drive locations, they can be sorted according to the distance from the user or the dealer rating. The third column can display multiple vehicle models of at least one car series corresponding to the vehicle type, and the multiple vehicle models are sorted according to the matching degree. The recommended vehicle content reflects the recommendation logic from the vehicle type to each vehicle model of the corresponding car series, and the vehicle model is displayed based on the matching degree of the vehicle model, which can enable users to understand the recommendation logic of the system, enhance users' trust in the recommendation system, and realize comprehensive and accurate vehicle content recommendation for users.

[0148] In some embodiments, in addition to the recommended vehicle model, the information displayed on the recommendation page may further display related explanations, such as:

[0149] "Based on your preferences, we recommend you to check out the following SUV models of 'Brand A', which all have excellent safety performance and space configuration and meet your needs."

[0150] "You may also like these models because they also belong to the 'Brand A' series, providing rich interior configurations and powerful power systems."

[0151] For example, the user has browsed "Brand A" SUVs many times in the past few weeks, focusing on their safety performance and interior configuration, especially their needs for family cars. The system identifies the following models in the "Brand A" SUV series that meet the user's preferences:

[0152] Brand A Model 1: High safety rating, spacious rear seat space, suitable for family use.

[0153] Brand A Model 2: Luxurious configuration and high-end interior, suitable for users who pursue comfort.

[0154] Brand A Model 3: Good fuel economy, suitable for long-distance driving.

[0155] Based on this, the system can recommend the following:

[0156] "We recommend the following 'Brand A' SUV models, all of which have excellent safety performance and space configuration and are very suitable for family use:"

[0157] Brand A Model 1: Safety score is 5 stars, rear seat space is spacious, suitable for family travel.

[0158] Brand A Model 2: Luxurious interior, equipped with the latest safety technology and ultimate comfort.

[0159] The system can also make recommendations based on the sales data of car series, giving priority to car series with better sales to increase the user's selection success rate.

[0160] In some embodiments, the vehicle content recommendation method of the embodiment of the present application may also include but is not limited to steps S701 to S702:

[0161] Step S701, updating user behavior data in response to a user's feedback operation on vehicle content;

[0162] Step S702, generating new vehicle content according to the updated user behavior data.

[0163] In this embodiment, please refer to Figure 3 After the recommended vehicle content is updated on the front-end page, the user will further generate interactive feedback operations, for example, the user does not click on the recommended vehicle type or clicks on the recommended vehicle type. This embodiment collects user feedback on the recommendation results and determines the user's satisfaction with the recommendation. The feedback data is extracted and normalized so that the data can be used for model optimization. The model is retrained based on user feedback, and the model parameters are updated to make it more suitable for user needs. Based on the updated model, new recommendation results are generated to form a virtuous circle.

[0164] Furthermore, the embodiment of the present application can obtain user feedback data and update the user's historical browsing data based on the user's real-time feedback data, thereby updating the performance indicator preference weight, so as to provide more accurate vehicle model recommendations later.

[0165] The real-time mechanism implemented in this paper can obtain the user's rating or feedback on the recommendation results in real time, and update the model or weight information in time according to these feedbacks to optimize the future recommendation effect. The key to this mechanism is to respond to user input in a timely manner, thereby realizing intelligent tuning, so that the system can continuously adapt to changes in user needs.

[0166] The embodiment of the present application also provides a vehicle content recommendation system, including:

[0167] The first module is used to obtain user feature data and user behavior data;

[0168] The second module is used to use deep learning technology to predict vehicle model preferences based on user feature data and user behavior data to obtain the vehicle type preferred by the user;

[0169] The third module is used to obtain user location information and determine the recommended test drive location based on the user location information and vehicle type;

[0170] The fourth module is used to obtain the vehicle model set corresponding to the vehicle type and the user's performance index preference weight information, and calculate the matching degree of each vehicle model in the vehicle model set according to the performance index preference weight information;

[0171] The fifth module is used to determine the recommended vehicle content according to the vehicle type, the recommended test drive location and the matching degree of each vehicle model in the vehicle model set.

[0172] It can be understood that the contents of the above-mentioned vehicle content recommendation method embodiment are all applicable to the present system embodiment, the functions specifically implemented by the present system embodiment are the same as those in the above-mentioned vehicle content recommendation method embodiment, and the beneficial effects achieved are also the same as those achieved by the above-mentioned vehicle content recommendation method embodiment.

[0173] The embodiment of the present application also provides an electronic device, the electronic device comprising: a memory, a processor, a program stored in the memory and executable on the processor, and a data bus for realizing connection and communication between the processor and the memory, and the program is executed by the processor to realize the above-mentioned vehicle content recommendation method. The electronic device can be any intelligent terminal including a tablet computer, a vehicle-mounted computer, etc.

[0174] See also Figure 4 , Figure 4 The hardware structure of an electronic device of another embodiment is illustrated, and the electronic device includes:

[0175] The processor 401 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application;

[0176] The memory 402 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 402 can store an operating system and other application programs. When the technical solution provided in the embodiment of this specification is implemented by software or firmware, the relevant program code is stored in the memory 402, and the processor 401 calls and executes the vehicle content recommendation method of the embodiment of this application;

[0177] Input / output interface 403, used to implement information input and output;

[0178] Communication interface 404, used to realize communication interaction between the device and other devices, which can be realized by wired mode (such as USB, network cable, etc.) or wireless mode (such as mobile network, WIFI, Bluetooth, etc.);

[0179] Bus 405 , which transmits information between various components of the device (e.g., processor 401 , memory 402 , input / output interface 403 , and communication interface 404 );

[0180] The processor 401 , the memory 402 , the input / output interface 403 and the communication interface 404 are connected to each other in communication within the device via the bus 405 .

[0181] An embodiment of the present application also provides a storage medium, which is a computer-readable storage medium used for computer-readable storage. The storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the above-mentioned vehicle content recommendation method.

[0182] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0183] The embodiments described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0184] Those skilled in the art will appreciate that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0185] The system embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment.

[0186] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0187] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0188] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0189] In the several embodiments provided in the present application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic. For example, the division of the above units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of systems or units, which can be electrical, mechanical or other forms.

[0190] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0191] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0192] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including multiple instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), disk or optical disk and other media that can store programs.

[0193] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but the scope of the rights of the present invention is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present invention should be within the scope of the rights of the present invention.

Claims

1. A vehicle content recommendation method, characterized in that: The following steps are involved: Obtain user feature data and user behavior data; Using deep learning technology, predicting vehicle type preference based on the user feature data and the user behavior data, and obtaining the vehicle type preferred by the user; Obtaining user location information, and determining a recommended test drive location based on the user location information and the vehicle type; Acquire a vehicle model set corresponding to the vehicle type and the user's performance indicator preference weight information, and calculate the matching degree of each vehicle model in the vehicle model set according to the performance indicator preference weight information; The recommended vehicle content is determined according to the vehicle type, the recommended test drive location, and the matching degree of each vehicle model in the vehicle model set.

2. The vehicle content recommendation method according to claim 1, characterized in that: The method of using deep learning technology to predict vehicle type preference based on the user feature data and the user behavior data to obtain the vehicle type preferred by the user includes the following steps: Preprocessing the user feature data and the user behavior data to obtain data to be analyzed; Extracting features of the data to be analyzed to obtain basic user features, browsing features, operation features, and comment content features; The format-converted user basic features, the browsing features, the operation features, and the comment content features are input into a user preference prediction model built based on deep learning technology to obtain the vehicle type preferred by the user.

3. The vehicle content recommendation method according to claim 1, characterized in that: The step of determining a recommended test drive location based on the user location information and the vehicle type includes the following steps: Obtaining a preset first test drive information table, wherein the first test drive information table includes vehicle inventory information of each test drive location; Filtering the test drive locations in the first test drive information table according to the user location information to obtain first candidate locations, and forming a second test drive information table according to the vehicle inventory information of each of the first candidate locations; Retrieving the vehicle inventory information in the second test drive information table according to the vehicle type to obtain a second candidate location where a vehicle corresponding to the vehicle type is stored; The second candidate location is determined as a recommended test drive location.

4. The vehicle content recommendation method according to claim 3, characterized in that: The step of screening the test drive locations in the first test drive information table according to the user location information to obtain a first candidate location includes the following steps: Calculating the distance between the user location and each test drive location in the first test drive information table according to the user location information and the location information of the test drive location; A test drive location whose distance is less than a first threshold is selected as a first candidate location.

5. The vehicle content recommendation method according to claim 1, characterized in that: The performance indicator preference weight information is obtained by the following steps: Get the user's historical browsing data for a preset time period before the current time; Extracting target content from the historical browsing data, wherein the target content includes semantic keywords related to performance indicators, browsed vehicle models, and dwell time of browsed vehicle models; Acquire performance characteristics of each browsed vehicle model, wherein the performance characteristics include at least one performance index; According to the stay time of the browsed vehicle model, the total time of each performance indicator in the performance characteristics of all the browsed vehicle models is counted, and the first weight of each performance indicator is determined according to the total time of each performance indicator; Counting the number of occurrences of the semantic keyword of each performance indicator in the target content, and determining a second weight of the performance indicator according to the number of occurrences; The user's performance indicator preference weight information is determined according to the first weight and the second weight of each performance indicator.

6. The vehicle content recommendation method according to claim 1, characterized in that: The step of calculating the matching degree of each vehicle model in the vehicle model set according to the performance indicator preference weight information comprises the following steps: Acquire performance characteristics of each vehicle model in the vehicle model set, wherein the performance characteristics include at least one performance indicator; determining a weight of each performance indicator in the performance characteristics of the vehicle model according to the performance indicator preference weight information; The matching degree of the vehicle model is determined according to the weights of all performance indicators in the performance characteristics of the vehicle model.

7. The vehicle content recommendation method according to claim 1, characterized in that: The vehicle content recommendation method further comprises the following steps: In response to a user's feedback operation on the vehicle content, updating the user behavior data; Generate new vehicle content based on the updated user behavior data.

8. A vehicle content recommendation system, characterized in that: include: The first module is used to obtain user feature data and user behavior data; The second module is used to use deep learning technology to predict vehicle type preference based on the user feature data and the user behavior data to obtain the vehicle type preferred by the user; The third module is used to obtain user location information and determine a recommended test drive location based on the user location information and the vehicle type; The fourth module is used to obtain the vehicle model set corresponding to the vehicle type and the user's performance indicator preference weight information, and calculate the matching degree of each vehicle model in the vehicle model set according to the performance indicator preference weight information; The fifth module is used to determine the recommended vehicle content according to the vehicle type, the recommended test drive location and the matching degree of each vehicle model in the vehicle model set.

9. An electronic device, characterized in that: The electronic device includes a memory, a processor, a program stored in the memory and executable on the processor, and a data bus for realizing connection and communication between the processor and the memory. When the program is executed by the processor, the steps of the method described in any one of claims 1 to 7 are realized.

10. A storage medium, the storage medium being a computer-readable storage medium, used for computer-readable storage, characterized in that: The storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of any one of claims 1 to 7.

Images