Data recommendation method and device based on artificial intelligence, computer equipment and medium

Abstract

The invention belongs to the technical field of artificial intelligence, and relates to a data recommendation method and device based on artificial intelligence, computer equipment and a storage medium, and the method comprises the steps: collecting input content and operation behavior data of a user in an interaction interface; performing appeal identification on the input content and the operation behavior data based on an identification rule to obtain a user type of the user; judging whether the user type is a vehicle type purchasing user or not; if yes, collecting vehicle category data input by the user and own condition information; acquiring historical related data and market dynamic information corresponding to the vehicle category data; performing recommendation processing on the vehicle category data, the own condition information, the historical related data and the market dynamic information based on a recommendation model to obtain a recommended vehicle model result; and returning the recommended vehicle model result to the user. In addition, the recommended vehicle model result can be stored in the block chain. The method can be applied to a data recommendation scene in the field of financial science and technology, and the efficiency, flexibility and accuracy of vehicle model recommendation processing are improved.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology and can be applied to the field of fintech, especially to data recommendation methods, devices, computer equipment and storage media based on artificial intelligence. Background Art

[0002] In the traditional automotive finance service model, the recommendation process for vehicle model selection mainly relies on preset simple rules, which results in low accuracy of recommended vehicle models. Specifically, traditional methods usually recommend vehicle models based on some basic information of users (such as budget range, vehicle type category preference), lacking in-depth analysis of users' multi-dimensional characteristics and needs. Such a rough recommendation method is difficult to accurately capture users' personalized needs, making the matching degree between the recommended vehicle models and users' actual needs insufficient, thereby reducing users' purchase intention and satisfaction.

[0003] For example, in the scenario of car loan purchase recommendation in the financial field, traditional methods may only recommend vehicle models based on users' budgets and preferred vehicle type categories (such as SUVs, sedans). However, this recommendation method ignores many key factors, such as users' credit status, driving habits, family travel needs, specific requirements for vehicle safety performance and technological configurations, etc. If a user has good credit, a long-term high-speed driving habit, a large family population in need of a large-space vehicle, and pays attention to vehicle safety performance and intelligent driving assistance functions, traditional methods may recommend a vehicle model with average space, low safety configuration and technological configuration just because of budget and vehicle type category matching, unable to meet users' actual needs in terms of driving safety, comfort and technological experience. Such recommendation deviation not only affects users' car purchase experience, but may also lead to a decline in the market competitiveness of financial institutions in automotive finance business, unable to effectively expand the customer base and increase business revenue.

[0004] In addition, traditional recommendation methods lack flexibility and adaptability when facing complex and changeable market environments and users' diverse needs. With the continuous development of the automotive market and the increasing personalization of users' needs, traditional recommendation models based on simple rules are difficult to meet actual needs. Therefore, traditional vehicle model recommendation methods have the problem of low recommendation accuracy. Summary of the Invention

[0005] The purpose of the embodiments of this application is to propose a data recommendation method, device, computer equipment and storage medium based on artificial intelligence to solve the technical problem of low recommendation accuracy existing in the existing vehicle model recommendation methods.

[0006] In a first aspect, a data recommendation method based on artificial intelligence is provided, including:

[0007] Collect input content and operation behavior data of users in the interaction interface;

[0008] Identify the user's type based on the preset recognition rules for the input content and the operation behavior data.

[0009] Determine whether the user type is a car-purchasing user.

[0010] If so, collect the vehicle category data and the user's own condition information input by the user.

[0011] Obtain the historical relevant data and market dynamics information corresponding to the vehicle category data.

[0012] Perform recommendation processing on the vehicle category data, the user's own condition information, the historical relevant data, and the market dynamics information based on the preset recommendation model to obtain the corresponding recommended vehicle model results.

[0013] Return the recommended vehicle model results to the user.

[0014] In a second aspect, a data recommendation device based on artificial intelligence is provided, including:

[0015] A first collection module for collecting the input content and operation behavior data of the user in the interaction interface.

[0016] An identification module for identifying the user's type based on the preset recognition rules for the input content and the operation behavior data.

[0017] A first judgment module for judging whether the user type is a car-purchasing user.

[0018] A second collection module for, if so, collecting the vehicle category data and the user's own condition information input by the user.

[0019] A first acquisition module for acquiring the historical relevant data and market dynamics information corresponding to the vehicle category data.

[0020] A recommendation module for performing recommendation processing on the vehicle category data, the user's own condition information, the historical relevant data, and the market dynamics information based on the preset recommendation model to obtain the corresponding recommended vehicle model results.

[0021] A return module for returning the recommended vehicle model results to the user.

[0022] In a third aspect, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above-mentioned data recommendation method based on artificial intelligence are implemented.

[0023] Fourthly, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned artificial intelligence-based data recommendation method are implemented.

[0024] In the solution implemented by the above-mentioned artificial intelligence-based data recommendation method, device, computer device, and storage medium, first, the input content and operation behavior data of the user in the interaction interface are collected; then, based on a preset recognition rule, the input content and the operation behavior data are subjected to appeal recognition to obtain the user type of the user; then, it is determined whether the user type is a vehicle purchase type user; if so, the vehicle category data and the user's own condition information input by the user are collected; subsequently, the historical relevant data and market dynamic information corresponding to the vehicle category data are obtained; further, based on a preset recommendation model, the vehicle category data, the user's own condition information, the historical relevant data, and the market dynamic information are subjected to recommendation processing to obtain a corresponding recommended vehicle model result; finally, the recommended vehicle model result is returned to the user. In this application, through the use of the recognition rule, the input content and operation behavior data of the user in the interaction interface are subjected to appeal recognition to obtain the user type of the user, and when it is detected that the user type is a vehicle purchase type user, the vehicle category data and the user's own condition information input by the user are collected, and the historical relevant data and market dynamic information corresponding to the vehicle category data are obtained. Furthermore, based on the use of the recommendation model, the vehicle category data, the user's own condition information, the historical relevant data, and the market dynamic information are subjected to recommendation processing to obtain a corresponding recommended vehicle model result, and the recommended vehicle model result is returned to the user. In this application, through the use of the recommendation model, the vehicle category data, the user's own condition information, the historical relevant data, and the market dynamic information are comprehensively analyzed to generate the final recommended vehicle model result, effectively improving the accuracy of the generated recommended vehicle model result, and improving the processing efficiency, processing flexibility, and processing accuracy of vehicle model recommendation. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] To more clearly illustrate the solutions in this application, the following will briefly introduce the drawings required for the description of the embodiments of this application. Obviously, the drawings in the following description are some embodiments of this application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0026] Figure 1 is an exemplary system architecture diagram to which this application can be applied;

[0027] Figure 2 is a flowchart of an embodiment of the artificial intelligence-based data recommendation method according to this application;

[0028] Figure 3It is a schematic structural diagram of an embodiment of an artificial intelligence-based data recommendation method according to the present application;

[0029] Figure 4 It is a schematic structural diagram of an embodiment of a computer device according to the present application. Detailed implementation manners

[0030] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs; the terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the specification and claims of this application and the above drawings are intended to cover non-exclusive inclusion. The terms "first", "second", etc. in the specification and claims of this application or the above drawings are used to distinguish different objects and not to describe a specific order.

[0031] Reference to "embodiment" herein means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of this application. The phrase does not necessarily refer to the same embodiment at all positions in the specification, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0032] In order to enable those skilled in the art of this technology to better understand the solution of this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the accompanying drawings.

[0033] As Figure 1 shown, the system architecture 100 may include a terminal device 101, a network 102, and a server 103. The terminal device 101 may be a laptop computer 1011, a tablet computer 1012, or a mobile phone 1013. The network 102 is a medium for providing a communication link between the terminal device 101 and the server 103. The network 102 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0034] Users can use the terminal device 101 to interact with the server 103 through the network 102 to receive or send messages, etc. Various communication client applications may be installed on the terminal device 101, such as a web browser application, a shopping application, a search application, an instant messaging tool, an email client, a social platform software, etc.

[0035] The terminal device 101 can be various electronic devices with a display screen and supporting web browsing. In addition to the laptop 1011, tablet computer 1012, or mobile phone 1013, the terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 (Moving Picture Experts Group Audio Layer IV) player, a laptop computer, a desktop computer, and so on.

[0036] The server 103 can be a server that provides various services, such as a background server that supports the pages displayed on the terminal device 101.

[0037] It should be noted that the artificial intelligence-based data recommendation method provided in the embodiments of the present application is generally executed by the server / terminal device. Correspondingly, the artificial intelligence-based data recommendation method is generally set in the server / terminal device.

[0038] It should be understood that Figure 1 the numbers of terminal devices, networks, and servers in

[0039] Continue to refer to Figure 2 , which shows a flowchart of an embodiment of the artificial intelligence-based data recommendation method according to the present application. According to different requirements, the order of the steps in this flowchart can be changed, and some steps can be omitted. The artificial intelligence-based data recommendation method provided in the embodiments of the present application can be applied to any scenario that requires data recommendation. Then, the artificial intelligence-based data recommendation method can be applied to the products in these scenarios. For example, vehicle model recommendation in the financial field. The artificial intelligence-based data recommendation method includes the following steps:

[0040] Step S201, collect the input content and operation behavior data of the user in the interaction interface.

[0041] In this embodiment, the electronic device on which the artificial intelligence-based data recommendation method runs (such as Figure 1The server / terminal device shown in the figure) can obtain input content and operation behavior data through a wired connection or a wireless connection. It should be noted that the above-mentioned wireless connection method may include but is not limited to 3G / 4G / 5G connection, WiFi connection, Bluetooth connection, WiMAX connection, Zigbee connection, UWB (ultra wi deband) connection, and other wireless connection methods currently known or to be developed in the future. The execution subject of this application is a data recommendation system, which can be referred to as a system. This application can be applied to business processing scenarios for recommending car models to users in the financial field. Among them, the above-mentioned interactive interface is an interactive interface in the system. When the user interacts through the interactive interface of the system, the system records the user's input content in real time, including the text of the question, the input car model data, etc. For example, if the user enters "A novice wants to know the characteristics of SUV models", the text will be stored in full. At the same time, various operation behavior data of the user on the interactive interface are recorded, such as the buttons clicked, the pages browsed, the dwell time, etc. For example, if the user clicks on different chapters of the car model knowledge popularization page multiple times, these operations will be recorded.

[0042] Step S202: performing demand recognition on the input content and the operation behavior data based on a preset recognition rule to obtain the user type of the user.

[0043] In this embodiment, the above identification rules are pre-constructed demand classification judgment rules. According to the rule content of the demand classification judgment rule, the demand identification of the input content and operation behavior data can be performed to obtain the user type of the user, and the user type includes a learning user and a vehicle purchase user. Specifically, the rule content of the demand classification judgment rule includes: Learning user judgment: When the user frequently consults vehicle model knowledge-related questions, such as vehicle model classification, technical parameter meaning, configuration function, etc., and the operation behavior is mainly concentrated on the knowledge learning page, it is judged as a learning user. For example, if the user consults more than 10 times in a month about the knowledge of the vehicle power system, it can be judged as a learning user. Vehicle purchase user judgment: If the user enters specific vehicle model data and own conditions, such as vehicle model brand, price range, usage scenario, etc., and explicitly requires vehicle model recommendation, it is judged as a vehicle purchase user. For example, if the user enters "I want to buy a joint venture SUV with a price between 200,000 and 300,000, mainly for urban commuting", it can be judged as a vehicle purchase user.

[0044] Step S203, determining whether the user type is a car model purchasing user.

[0045] In this embodiment, the user type may be identified by content recognition to determine whether the user type is a vehicle-buying user, wherein the user type includes a learning user or a vehicle-buying user.

[0046] Step S204, if so, collect the vehicle category data and the user's own condition information input by the user.

[0047] In this embodiment, if the user type is a vehicle purchase user, it can be detected whether the input content of the user in the interaction interface includes vehicle category data and the user's own condition information. If it includes, directly extract the vehicle category data and the user's own condition information from the above input content. If it does not include, it can be obtained by means of detailed prompts or information collection forms. Specifically, the detailed prompt method includes: when the user does not know how to choose among many vehicle models, the model will pop up a prompt box to guide the user to provide more detailed vehicle category data (such as SUV, sedan, MPV, etc.) and the user's own basic conditions (such as budget, usage scenario, preferred configuration, etc.). For example, the prompt box shows "To recommend vehicle models for you more accurately, please inform the vehicle category you like (such as SUV) and the budget range (such as 100,000 - 150,000)". The information collection form method includes: providing a convenient information collection form for the user to easily input relevant information. The form can include various input methods such as drop-down menus, sliders, and text boxes to improve the user input efficiency.

[0048] Step S205, obtain the historical relevant data and market dynamic information corresponding to the vehicle category data.

[0049] In this embodiment, the historical relevant data and market dynamic information corresponding to the above vehicle category data can be collected from a preset data source. Among them, the above historical relevant data includes data such as the sales situation and user evaluations of the vehicle models corresponding to the above vehicle category data; the market dynamic information includes data such as price fluctuations, promotional activities, and new vehicle model releases of the vehicle models corresponding to the above vehicle category data. In addition, the above data source can include channels such as automobile manufacturers, dealers, automobile media, automobile websites, forums, and social media.

[0050] Step S206, perform a recommendation process on the vehicle category data, the user's own condition information, the historical relevant data, and the market dynamic information based on a preset recommendation model to obtain a corresponding recommended vehicle model result.

[0051] In this embodiment, the vehicle category data, the user's own condition information, the historical relevant data, and the market dynamic information can be integrated to obtain integrated data and input it into the above recommendation model. The recommendation model will analyze the input integrated data and output vehicle models that meet the conditions, which are used as the corresponding recommended vehicle model results. Among them, for the specific construction process of the above recommendation model, this application will further describe the details in subsequent specific embodiments and will not elaborate too much here.

[0052] Step S207, return the recommended vehicle model result to the user.

[0053] In this embodiment, for the specific implementation process of returning the recommended vehicle model result to the user, this application will further describe the details thereof in subsequent specific embodiments and will not elaborate too much here.

[0054] This application first collects the input content and operation behavior data of the user in the interaction interface; then performs requirement recognition on the input content and the operation behavior data based on preset recognition rules to obtain the user type of the user; then determines whether the user type is a vehicle purchase type user; if so, collects the vehicle category data and self-condition information input by the user; subsequently obtains the historical relevant data and market dynamic information corresponding to the vehicle category data; further performs recommendation processing on the vehicle category data, the self-condition information, the historical relevant data, and the market dynamic information based on a preset recommendation model to obtain a corresponding recommended vehicle model result; finally returns the recommended vehicle model result to the user. This application performs requirement recognition on the input content and operation behavior data of the user in the interaction interface based on the use of recognition rules to obtain the user type of the user, and when detecting that the user type is a vehicle purchase type user, will collect the vehicle category data and self-condition information input by the user, and obtain the historical relevant data and market dynamic information corresponding to the vehicle category data, and then perform recommendation processing on the vehicle category data, self-condition information, historical relevant data, and market dynamic information based on the use of the recommendation model to obtain a corresponding recommended vehicle model result, and returns the recommended vehicle model result to the user. This application comprehensively analyzes the vehicle category data, self-condition information, historical relevant data, and market dynamic information based on the use of the recommendation model to generate the final recommended vehicle model result, effectively improving the accuracy of the generated recommended vehicle model result, and improving the processing efficiency, processing flexibility, and processing accuracy of vehicle model recommendation.

[0055] In some optional implementation manners, before step S206, the above electronic device may further perform the following steps:

[0056] Obtain initial vehicle model data from a preset data source.

[0057] In this embodiment, the above-mentioned preset data sources may at least include channels such as automobile manufacturers, dealerships, automotive media, automotive websites, forums, and social media. By establishing cooperative relationships with automobile manufacturers, dealerships, automotive media, etc., detailed data of various models on the market can be obtained. For example, obtain the basic parameters, configuration information, price ranges, etc. of vehicle models from automobile manufacturers; obtain sales records and promotional activity information from dealerships; obtain market research reports and user evaluations from automotive media. And use web crawler technology to collect discussions and evaluations of users on different vehicle models from major automotive websites, forums, social media and other channels. A crawler program can be written using Python's Scrapy framework, setting appropriate crawling rules to extract points that users are concerned about, such as the advantages, disadvantages, and usage experiences of vehicle models. Subsequently, all the obtained data will be integrated and processed to obtain the corresponding initial vehicle model data.

[0058] Construct sample data based on the initial vehicle model data.

[0059] In this embodiment, the corresponding sample data can be constructed by performing data cleaning processing, data standardization processing, and feature engineering processing on the obtained initial vehicle model data. Among them, data cleaning processing includes: cleaning the collected data to remove duplicate, incorrect, and incomplete data. For example, use a data deduplication algorithm to delete duplicate vehicle model records; for records lacking key information (such as price, configuration), you can choose to delete them or fill them according to other relevant data. Data standardization processing includes: converting data in different formats into a unified format. For example, unify the data of vehicle model sizes into length, width, and height values in millimeters; unify the date format into the standard YYYY-MM-DD format. Feature engineering processing includes: extracting features that have an important impact on vehicle model decision-making and performing encoding and conversion. For example: Safety performance features: Features such as crash test ratings and the number of airbags can be extracted and encoded into numerical forms. Fuel economy features: Extract the fuel consumption per 100 kilometers data and perform standardization processing. Technology configuration features: Encode the levels of intelligent driving assistance systems, the presence or absence of vehicle networking functions, etc.

[0060] Call the pre-constructed initial large model.

[0061] In this embodiment, for the construction process of the above-mentioned initial large model, this application will further describe the details in subsequent specific embodiments and will not elaborate too much here.

[0062] Based on the preset model training strategy, use the sample data to train the initial large model to obtain the corresponding first large model.

[0063] In this embodiment, the above model training strategy can be generated by combining various training methods such as supervised learning, unsupervised learning, and reinforcement learning, and this model training strategy is used as the training method for the initial large model. Then, according to this training method, after dividing the above sample data into a training set, a validation set, and a test set, the training set is used to train the initial large model, and the parameters of the model are adjusted through the backpropagation algorithm to make the output result of the model as close as possible to the true label. During the training process, the validation set is used to monitor the performance of the model to prevent overfitting. After the training is completed, the test set is used to evaluate the generalization ability of the model, thereby obtaining the trained first large model.

[0064] Perform a performance evaluation on the first large model based on a preset evaluation metric to obtain the corresponding performance evaluation result.

[0065] In this embodiment, there is no specific limitation on the selection of the above evaluation metrics. Appropriate evaluation metrics can be selected to evaluate the performance of the model, such as accuracy, recall rate, F1 value, etc. For the vehicle model recommendation scenario, the recommendation accuracy (the proportion of the recommended vehicle models that meet the user's needs) and user satisfaction can be used as evaluation metrics. Among them, the performance evaluation of the first large model can be performed using the above test set to obtain the performance evaluation result corresponding to this evaluation metric.

[0066] Perform model optimization processing on the first large model based on the performance evaluation result to obtain the optimized second large model.

[0067] In this embodiment, by optimizing the above first large model according to the obtained performance evaluation result, specifically, the performance of the model can be improved by adjusting the hyperparameters of the model (such as the learning rate, the number of neurons in the hidden layer), improving the feature engineering method, increasing the training data, etc., and then the optimized second large model is obtained and used as the final recommendation model.

[0068] Use the second large model as the recommendation model.

[0069] In this embodiment, the obtained recommendation model can also be updated correspondingly according to a preset model update strategy. Among them, the model update strategy includes a real-time data update strategy and a personalized update strategy. The real-time data update strategy includes: establishing a real-time data update mechanism, and automatically obtaining the latest vehicle model data and user feedback information through continuous connection with the data source. For example, using a message queue (such as Kafka) to receive new data, and triggering the model update process when new data arrives. The personalized update strategy includes: setting personalized data update strategies according to the concerns of different user groups. For example: for the young user group, focus on the technology configuration and appearance design data of the vehicle model. When a new vehicle model is released or the technology configuration of an existing vehicle model is upgraded, update the relevant data of the model in a timely manner. For family users, focus on the space and safety data of the vehicle model. When new safety test results are announced or the space data of the vehicle model changes, update the model.

[0070] This application obtains initial vehicle model data from a preset data source, and constructs sample data based on the initial vehicle model data; then calls a pre-constructed initial large model; then, based on a preset model training strategy, uses the sample data to train the initial large model to obtain a corresponding first large model; subsequently, based on a preset evaluation index, performs performance evaluation on the first large model to obtain a corresponding performance evaluation result; further, based on the performance evaluation result, performs model optimization processing on the first large model to obtain an optimized second large model; finally, uses the second large model as the recommendation model. This application obtains initial vehicle model data from a preset data source, constructs sample data based on the initial vehicle model data, then based on the model training strategy, uses the sample data to train the initial large model to obtain the first large model, then performs performance evaluation on the first large model based on the use of the evaluation index to obtain the performance evaluation result, and further performs model optimization processing on the first large model based on the use of the performance evaluation result to obtain the optimized second large model and use it as the recommendation model, so as to efficiently and accurately construct a recommendation model that can process various types of data and provide support for vehicle model decision-making, improve the construction efficiency of the recommendation model, and ensure the model efficiency of the obtained recommendation model.

[0071] In some alternative implementation manners of this embodiment, before the step of calling the pre-constructed initial large model, the above electronic device may further perform the following steps:

[0072] Obtain a preset distributed architecture, and generate a specified model based on the distributed architecture.

[0073] In this embodiment, the selection of the above-mentioned distributed architecture is not specifically limited and can be determined according to actual business usage requirements. For example, a deep learning framework can be utilized. Among them, the distributed architecture helps to process large-scale data and accelerate model training. For example, a computing cluster is composed of multiple servers, and each server is responsible for processing part of the data and computing tasks. In addition, the initial large model of the recommended model to be generated can be constructed by utilizing the distributed architecture.

[0074] An input layer including a first preset number of input channels is set in the specified model to obtain a corresponding first generation model.

[0075] In this embodiment, by setting multiple input channels in the above-mentioned specified model to receive different types of data. Among them, the selection of the above-mentioned first preset number is not specifically limited and can be determined according to actual business requirements. For example, the first preset number can be set to 3, and the input channels include: a basic vehicle model attribute data channel: receiving information such as vehicle model size (length, width, height), type (SUV, sedan, MPV, etc.). A high-tech function data channel: receiving data such as intelligent driving assistance systems and vehicle networking functions. A pre-sale, in-sale, and after-sale data channel: receiving market research reports, sales records, maintenance records, etc.

[0076] A second preset number of hidden layers is set in the first generation model to obtain a corresponding second generation model.

[0077] In this embodiment, by designing multiple hidden layers in the above-mentioned second generation model, each hidden layer contains multiple neurons. And activation functions (such as ReLU, Sigmoid) are used to introduce non-linear factors to enhance the model's ability to capture complex relationships in the data. Among them, the selection of the above-mentioned second preset number is not specifically limited and can be determined according to actual business requirements. For example, 3-5 hidden layers can be set, and each hidden layer contains 100-500 neurons.

[0078] Obtain a preset application scenario.

[0079] In this embodiment, the above-mentioned application scenario includes at least a vehicle model recommendation scenario and a knowledge answering scenario.

[0080] An output layer corresponding to the application scenario is set in the second generation model to obtain a corresponding third generation model.

[0081] In this embodiment, by designing a corresponding output layer according to different application scenarios in the above-mentioned second generation model. Specifically, for the vehicle model recommendation scenario, the output layer can output a list of recommended vehicle models and recommendation scores; for the knowledge answering scenario, the output layer can output relevant vehicle model knowledge texts.

[0082] Use the third generation model as the initial large model.

[0083] In this application, a preset distributed architecture is obtained, and a specified model is generated based on the distributed architecture; then, an input layer including a first preset number of input channels is set in the specified model to obtain a corresponding first generation model; after that, a second preset number of hidden layers are set in the first generation model to obtain a corresponding second generation model; subsequently, a preset application scenario is obtained, and an output layer corresponding to the application scenario is set in the second generation model to obtain a corresponding third generation model; finally, the third generation model is used as the initial large model. In this application, by obtaining a preset distributed architecture and generating a specified model based on the distributed architecture, and then by setting an input layer including a first preset number of input channels, a second preset number of hidden layers, and an output layer corresponding to a preset application scenario in the specified model, an initial large model that meets the requirements can be constructed efficiently and accurately, effectively ensuring the adaptability and accuracy of the generated initial large model.

[0084] In some alternative implementation manners, before the step of training the initial large model with the sample data based on a preset model training strategy to obtain a corresponding first large model, the above electronic device may further perform the following steps:

[0085] Obtain a preset supervised learning training strategy, an unsupervised learning training strategy, and a reinforcement learning training strategy.

[0086] In this embodiment, the strategy content of the above supervised learning training strategy includes: using labeled data (such as car model recommendation cases) to train the model to enable the model to learn the mapping relationship between the input data and the output result. The strategy content of the above unsupervised learning training strategy includes: performing clustering analysis on unlabeled data (such as user discussion texts) to discover potential patterns in the data and the concerns of user groups. The strategy content of the above reinforcement learning training strategy includes: through interaction with the user, continuously optimizing the model's recommendation strategy according to the user's feedback.

[0087] Perform a combination process on the supervised learning training strategy, the unsupervised learning training strategy, and the reinforcement learning training strategy to obtain a corresponding combined training strategy.

[0088] In this embodiment, by performing a combination process on the above supervised learning training strategy, unsupervised learning training strategy, and reinforcement learning training strategy, the model can learn the relationships and rules between different types of data, and a corresponding combined training strategy can be obtained.

[0089] Use the combined training strategy as the model training strategy.

[0090] This application obtains a preset supervised learning training strategy, an unsupervised learning training strategy, and a reinforcement learning training strategy; then combines and processes the supervised learning training strategy, the unsupervised learning training strategy, and the reinforcement learning training strategy to obtain a corresponding combined training strategy; subsequently, uses the combined training strategy as the model training strategy. By obtaining the supervised learning training strategy, the unsupervised learning training strategy, and the reinforcement learning training strategy, and then combining and processing the supervised learning training strategy, the unsupervised learning training strategy, and the reinforcement learning training strategy, this application can intelligently and accurately construct the required model training strategy, improving the intelligence and accuracy of the generation of the model training strategy. This is beneficial for subsequent training of the initial large model using the constructed model training strategy, and can effectively improve the model effect of the recommended model generated by the training.

[0091] In some optional implementation manners, step S207 includes the following steps:

[0092] Perform a scoring process on all the recommended vehicle model results to obtain corresponding scoring results.

[0093] In this embodiment, communicate with experts and senior users in the automotive field in advance, and combine market research data to determine the key factors affecting vehicle model selection, such as price, performance, configuration, word-of-mouth, etc. Then score the recommended vehicle model results selected according to the key factors. Specifically, use the analytic hierarchy process to set different weights for each factor, and adopt the weighted summation algorithm to calculate the comprehensive score according to the performance of the vehicle model in each factor to obtain the scoring results of the recommended vehicle model results.

[0094] Perform a sorting process on all the recommended vehicle model results based on the scoring results to obtain a corresponding recommended vehicle model list.

[0095] In this embodiment, all the recommended vehicle model results can be sorted according to the numerical values of the scoring results from largest to smallest to obtain a sorted recommended vehicle model list.

[0096] Perform a supplementary process on the decision-making basis of the recommended vehicle model list to obtain a corresponding target recommended vehicle model list.

[0097] In this embodiment, the supplementary process of the decision-making basis for the above-mentioned recommended vehicle model list can be completed by providing detailed decision-making bases and suggestions for each recommended vehicle model result included in the recommended vehicle model list, such as the advantages (such as large space, low fuel consumption) and disadvantages (such as low brand awareness) of the vehicle model, and the comparison with other vehicle models. For example, when recommending a compact SUV, explain its advantages in terms of space utilization and fuel economy compared with vehicles in the same class.

[0098] Obtain a preset target feedback method.

[0099] In this embodiment, the selection of the above-mentioned target feedback method is not specifically limited. For example, it may include any one of methods such as email sending, interface display, message reminder, etc.

[0100] Based on the target feedback method, the target recommended vehicle model list is returned to the user.

[0101] In this embodiment, the target recommended vehicle model list can be returned to the user according to the selected target feedback method. Among them, the user can be allowed to give feedback on the recommended vehicle model list, such as expressing satisfaction, dissatisfaction, or putting forward further requirements. The system will adjust the recommendation plan in a timely manner according to the user's feedback to provide results that better meet the user's needs.

[0102] In this application, all the recommended vehicle model results are scored to obtain corresponding scoring results; then, based on the scoring results, all the recommended vehicle model results are sorted to obtain a corresponding recommended vehicle model list; after that, supplementary processing of decision-making basis is performed on the recommended vehicle model list to obtain a corresponding target recommended vehicle model list; subsequently, a preset target feedback method is obtained; finally, based on the target feedback method, the target recommended vehicle model list is returned to the user. After generating the recommended vehicle model results based on the use of the recommendation model, this application obtains scoring results by scoring the recommended vehicle model results, sorts all the recommended vehicle model results based on the scoring results to obtain a recommended vehicle model list, and then performs supplementary processing of decision-making basis on the recommended vehicle model list to obtain a target recommended vehicle model list, effectively improving the intelligence, accuracy, and richness of the generated target recommended vehicle model list. Subsequently, by using the target feedback method to return the target recommended vehicle model list to the user, a more objective and accurate recommended vehicle model plan is accurately provided for the user, improving the push intelligence and accuracy of the recommended vehicle model plan, as well as improving the user's usage experience.

[0103] In some alternative implementation manners of this embodiment, after step S203, the above-mentioned electronic device may further perform the following steps:

[0104] If the user type is not a vehicle purchase type user, it is determined whether the user type is a learning type user.

[0105] In this embodiment, the content of the above-mentioned user type may include a vehicle purchase type user or a learning type user.

[0106] If so, obtain the consultation question type from the input content.

[0107] In this embodiment, the corresponding consultation question type can be obtained by extracting consultation questions from the input content of the user in the interaction interface.

[0108] Obtain learning resources corresponding to the type of the consultation question from a preset vehicle model knowledge base.

[0109] In this embodiment, the above vehicle model knowledge base is a pre-constructed database containing vehicle model knowledge of various vehicle models. Relevant learning resources can be matched from the above vehicle model knowledge base according to the type of questions consulted by the user. For example, if the user consults about the charging problem of new energy vehicles, the system will recommend relevant articles and videos such as the charging principle of new energy vehicles, the types of charging piles, and charging precautions. In addition, personalized learning content can be recommended in combination with the user's historical learning records and interest preferences. For example, if the user was previously interested in luxury brand vehicle models, the system will give priority to recommending learning resources on the technical characteristics and historical development of luxury brand vehicle models.

[0110] Construct a learning path corresponding to the learning resources.

[0111] In this embodiment, the vehicle model knowledge included in the above learning resources can be classified according to difficulty levels, such as elementary (basic concepts of vehicle models), intermediate (technical parameters of vehicle models), and advanced (design and R & D of vehicle models). Then, according to the user's learning level and progress, a suitable learning path can be planned for the user. In addition, clear goals can be set for each learning stage. For example, the goal of the elementary stage is to enable the user to understand the basic classification of vehicle models and common terms. The system will guide the user to gradually complete the stage goals according to the user's learning situation.

[0112] Send the learning resources and the learning path to the user.

[0113] In this embodiment, the above learning resources and learning path can be sent to the user by means of interface display. Among them, the user's learning progress can also be recorded, including the content already learned, learning duration, etc. For example, the system can display that the user has completed 80% of the learning content of the "Vehicle Exterior Design" chapter. And learning feedback can be provided to the user regularly, such as a summary of learning achievements and an evaluation of knowledge mastery. At the same time, a learning incentive mechanism, such as learning points, medals, etc., is set up to encourage the user to continue learning.

[0114] If it is detected that the user type is not a car - purchasing user in this application, then it is determined whether the user type is a learning user; if so, the consultation question type is obtained from the input content; then learning resources corresponding to the consultation question type are obtained from a preset vehicle model knowledge base; then a learning path corresponding to the learning resources is constructed; subsequently, the learning resources and the learning path are sent to the user. When it is detected that the user type of the user is a learning user in this application, by obtaining the consultation question type from the input content, then automatically and intelligently obtaining learning resources corresponding to the consultation question type based on the use of the vehicle model knowledge base, and constructing a learning path corresponding to the learning resources, and then sending the learning resources and the learning path to the user, thus intelligently and accurately realizing different recommendation results according to different user demands, quickly and effectively improving the learning user's understanding of vehicle models, improving the intelligence of data recommendation, and improving the user experience.

[0115] In some alternative implementation manners of this embodiment, after step S206, the above - mentioned electronic device may further perform the following steps:

[0116] Extract key index data from the recommended vehicle model results based on a preset index type.

[0117] In this embodiment, the above - mentioned index type may specifically include price type, fuel consumption type, power parameter type, etc. Key index data matching the above - mentioned index type can be extracted from the above - mentioned recommended vehicle model results, including, for example, price, fuel consumption, power parameters, etc.

[0118] Call a preset chart - drawing tool.

[0119] In this embodiment, the above - mentioned chart - drawing tool is a tool with chart - drawing function.

[0120] Perform chart drawing on the key index data based on the chart - drawing tool to obtain a corresponding first chart.

[0121] In this embodiment, by using the chart - drawing tool, an intuitive chart can be generated according to the extracted data. For example, a bar chart is used to compare the prices and fuel consumptions of recommended vehicle models, and a line chart is used to show the power performance curve of vehicle models.

[0122] Perform text - explanation processing on the first chart based on the recommended vehicle model results to obtain a corresponding second chart.

[0123] In this embodiment, by collecting pictures of the appearance and interior of the recommended vehicle models corresponding to the above - mentioned recommended vehicle model results and integrating them with the chart and text description, the characteristics, advantages, and applicable scenarios of the vehicle models can be described in detail in words, making the display content more rich and vivid, and obtaining a corresponding second chart.

[0124] Perform a display process on the second chart.

[0125] In this embodiment, the display process of the second chart can be completed by calling a pre-built vehicle model recommendation interface and presenting the second chart in the vehicle model recommendation interface.

[0126] In this application, key indicator data is extracted from the recommended vehicle model results based on a preset indicator type; then a preset chart drawing tool is called; then a chart is drawn for the key indicator data based on the chart drawing tool to obtain a corresponding first chart; subsequently, a text explanation process is performed on the first chart based on the recommended vehicle model results to obtain a corresponding second chart; finally, a display process is performed on the second chart. After generating the recommended vehicle model results based on the use of the recommendation model, this application will extract key indicator data from the recommended vehicle model results based on the use of a preset indicator type, then draw a chart for the key indicator data based on the use of the chart drawing tool to obtain a first chart, and then perform a text explanation process on the first chart based on the recommended vehicle model results to obtain a second chart, and subsequently perform a display process on the second chart, which is beneficial to helping users intuitively understand the key indicators of the recommendation results and make more informed choices, effectively improving the user experience.

[0127] In some alternative implementation manners, the obtained user information has obtained the consent of the user and complies with the provisions of relevant laws and relevant policies.

[0128] In addition, the non-company software tools or components that appear in the embodiments of this application are only introduced by way of example and do not represent actual use.

[0129] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0130] It should be emphasized that to further ensure the privacy and security of the above recommended vehicle model results, the above recommended vehicle model results can also be stored in a node of a blockchain.

[0131] The blockchain referred to in this application is a new application mode of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, and encryption algorithms. A blockchain, in essence, is a decentralized database, a string of data blocks generated by using cryptographic methods. Each data block contains information about a batch of network transactions, which is used to verify the validity (anti-counterfeiting) of the information and generate the next block. A blockchain can include a blockchain underlying platform, a platform product service layer, and an application service layer, etc.

[0132] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, sense the environment, acquire knowledge, and use knowledge to obtain the best results in theory, methods, technologies, and application systems.

[0133] Artificial intelligence basic technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation / interaction systems, and mechatronics. Artificial intelligence software technologies mainly include several major directions such as computer vision technology, robotics, biometric technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0134] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through computer-readable instructions. The computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, read-only memory (ROM), or a random access memory (RAM), etc.

[0135] It should be understood that although the steps in the flowchart of the accompanying drawings are shown in sequence according to the indication of the arrows, these steps do not necessarily have to be executed in the order indicated by the arrows. Unless there is a clear indication in this article, the execution of these steps has no strict order restriction, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages. These sub-steps or stages do not necessarily have to be completed at the same moment, but can be executed at different moments, and their execution order does not necessarily have to be sequential, but can be executed alternately or in turn with at least a part of other steps or sub-steps or stages of other steps.

[0136] Further reference Figure 3 to Figure 2 As an implementation of the method shown above, the present application provides an embodiment of a data recommendation method based on artificial intelligence. This device embodiment corresponds to the method embodiment shown in Figure 2 and this device can be specifically applied to various electronic devices.

[0137] Such as Figure 3As shown in the figure, the artificial intelligence-based data recommendation device 300 described in this embodiment includes: a first collection module 301, an identification module 302, a first judgment module 303, a second collection module 304, a first acquisition module 305, a recommendation module 306, and a return module 307.

[0138] Among them:

[0139] The first collection module 301 is used to collect the input content and operation behavior data of the user in the interaction interface;

[0140] The identification module 302 is used to perform claim identification on the input content and the operation behavior data based on a preset identification rule to obtain the user type of the user;

[0141] The first judgment module 303 is used to judge whether the user type is a car purchase type user;

[0142] The second collection module 304 is used to, if so, collect the vehicle category data and its own condition information input by the user;

[0143] The first acquisition module 305 is used to acquire the historical relevant data and market dynamic information corresponding to the vehicle category data;

[0144] The recommendation module 306 is used to perform recommendation processing on the vehicle category data, the own condition information, the historical relevant data, and the market dynamic information based on a preset recommendation model to obtain a corresponding recommended vehicle model result;

[0145] The return module 307 is used to return the recommended vehicle model result to the user.

[0146] In this embodiment, the operations respectively performed by the above modules or units correspond one by one to the steps of the artificial intelligence-based data recommendation method in the foregoing embodiment, and will not be elaborated here.

[0147] In some optional implementation manners of this embodiment, the artificial intelligence-based data recommendation device further includes:

[0148] A second acquisition module, which is used to acquire initial vehicle model data from a preset data source;

[0149] A first construction module, which is used to construct sample data based on the initial vehicle model data;

[0150] A first call module, which is used to call a pre-constructed initial large model;

[0151] A training module, which is used to train the initial large model with the sample data based on a preset model training strategy to obtain a corresponding first large model;

[0152] An evaluation module for performing performance evaluation on the first large model based on preset evaluation metrics to obtain corresponding performance evaluation results;

[0153] An optimization module for performing model optimization processing on the first large model based on the performance evaluation results to obtain an optimized second large model;

[0154] A first determination module for using the second large model as the recommendation model.

[0155] In some alternative implementation manners of this embodiment, the data recommendation device based on artificial intelligence further includes:

[0156] A generation module for obtaining a preset distributed architecture and generating a specified model based on the distributed architecture;

[0157] A first setting module for setting an input layer including a first preset number of input channels in the specified model to obtain a corresponding first generated model;

[0158] A second setting module for setting a second preset number of hidden layers in the first generated model to obtain a corresponding second generated model;

[0159] A third acquisition module for obtaining a preset application scenario;

[0160] A third setting module for setting an output layer corresponding to the application scenario in the second generated model to obtain a corresponding third generated model;

[0161] A second determination module for using the third generated model as the initial large model.

[0162] In some alternative implementation manners of this embodiment, the data recommendation device based on artificial intelligence further includes:

[0163] A fourth acquisition module for obtaining a preset supervised learning training strategy, an unsupervised learning training strategy, and a reinforcement learning training strategy;

[0164] A combination module for performing combination processing on the supervised learning training strategy, the unsupervised learning training strategy, and the reinforcement learning training strategy to obtain a corresponding combined training strategy;

[0165] A third determination module for using the combined training strategy as the model training strategy.

[0166] In some alternative implementation manners of this embodiment, the return module 307 includes:

[0167] A scoring sub-module for performing scoring processing on all the recommended vehicle model results to obtain corresponding scoring results;

[0168] A sorting sub-module, configured to sort all the recommended vehicle model results based on the scoring results to obtain a corresponding list of recommended vehicle models;

[0169] A supplement sub-module, configured to supplement the decision-making basis of the list of recommended vehicle models to obtain a corresponding list of target recommended vehicle models;

[0170] An acquisition sub-module, configured to acquire a preset target feedback method;

[0171] A return sub-module, configured to return the list of target recommended vehicle models to the user based on the target feedback method.

[0172] In some alternative implementation manners of this embodiment, the data recommendation device based on artificial intelligence further includes:

[0173] A second judgment module, configured to judge whether the user type is a learning user if the user type is not a vehicle purchase user;

[0174] A fifth acquisition module, configured to, if so, acquire the consultation question type from the input content;

[0175] A sixth acquisition module, configured to acquire learning resources corresponding to the consultation question type from a preset vehicle model knowledge base;

[0176] A second construction module, configured to construct a learning path corresponding to the learning resources;

[0177] A sending module, configured to send the learning resources and the learning path to the user.

[0178] In some alternative implementation manners of this embodiment, the data recommendation device based on artificial intelligence further includes:

[0179] An extraction module, configured to extract key index data from the recommended vehicle model results based on a preset index type;

[0180] A second calling module, configured to call a preset chart drawing tool;

[0181] A drawing module, configured to draw a chart for the key index data based on the chart drawing tool to obtain a corresponding first chart;

[0182] A processing module, configured to perform literal interpretation processing on the first chart based on the recommended vehicle model results to obtain a corresponding second chart;

[0183] A display module, configured to perform display processing on the second chart.

[0184] To solve the above technical problems, the embodiments of the present application also provide a computer device. For details, please refer to Figure 4 , Figure 4 , which is the basic structural block diagram of the computer device in this embodiment.

[0185] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are communicatively connected to each other through a system bus. It should be noted that only the computer device 4 with components 41-43 is shown in the figure. However, it should be understood that it is not required to implement all the shown components, and more or fewer components can be alternatively implemented. Among them, those skilled in the art of the present technology can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0186] The computer device can be a desktop computer, a notebook, a palm computer, a cloud server, or other computing devices. The computer device can interact with the user through a keyboard, a mouse, a remote control, a touchpad, a voice control device, or other means.

[0187] The memory 41 at least includes one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (such as SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 41 may be an internal storage unit of the computer device 4, such as the hard disk or memory of the computer device 4. In other embodiments, the memory 41 may also be an external storage device of the computer device 4, such as a plug-in hard disk equipped on the computer device 4, a Smart Media Card (SMC), a Secure Digital (SD) card, a FlashCard, etc. Of course, the memory 41 may also include both the internal storage unit and the external storage device of the computer device 4. In this embodiment, the memory 41 is generally used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions of the data recommendation method based on artificial intelligence. In addition, the memory 41 may also be used to temporarily store various types of data that have been output or will be output.

[0188] In some embodiments, the processor 42 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips. The processor 42 is generally used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to run the computer-readable instructions stored in the memory 41 or process data, such as running the computer-readable instructions of the data recommendation method based on artificial intelligence.

[0189] The network interface 43 may include a wireless network interface or a wired network interface, and the network interface 43 is generally used to establish a communication connection between the computer device 4 and other electronic devices.

[0190] Compared with the prior art, the embodiments of the present application mainly have the following beneficial effects:

[0191] In the embodiments of the present application, the present application identifies the user's requests for the input content and operation behavior data of the user in the interaction interface based on the use of recognition rules, obtains the user type of the user, and when it is detected that the user type is a car purchase user type, it will collect the vehicle category data and its own condition information input by the user, and obtain the historical relevant data and market dynamics information corresponding to the vehicle category data. Furthermore, based on the use of the recommendation model, it performs recommendation processing on the vehicle category data, its own condition information, historical relevant data, and market dynamics information to obtain the corresponding recommended vehicle model results, and returns the recommended vehicle model results to the user. The present application comprehensively analyzes the vehicle category data, its own condition information, historical relevant data, and market dynamics information based on the use of the recommendation model to generate the final recommended vehicle model results, effectively improving the accuracy of the generated recommended vehicle model results, and improving the processing efficiency, processing flexibility, and processing accuracy of vehicle model recommendation.

[0192] The present application also provides another implementation manner, that is, to provide a computer-readable storage medium, the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor, so that the at least one processor executes the steps of the data recommendation method based on artificial intelligence as described above.

[0193] Compared with the prior art, the embodiments of the present application mainly have the following beneficial effects:

[0194] In the embodiments of the present application, the present application identifies the user's requests for the input content and operation behavior data of the user in the interaction interface based on the use of recognition rules, obtains the user type of the user, and when it is detected that the user type is a car purchase user type, it will collect the vehicle category data and its own condition information input by the user, and obtain the historical relevant data and market dynamics information corresponding to the vehicle category data. Furthermore, based on the use of the recommendation model, it performs recommendation processing on the vehicle category data, its own condition information, historical relevant data, and market dynamics information to obtain the corresponding recommended vehicle model results, and returns the recommended vehicle model results to the user. The present application comprehensively analyzes the vehicle category data, its own condition information, historical relevant data, and market dynamics information based on the use of the recommendation model to generate the final recommended vehicle model results, effectively improving the accuracy of the generated recommended vehicle model results, and improving the processing efficiency, processing flexibility, and processing accuracy of vehicle model recommendation.

[0195] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-described embodiment methods can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases, the former is a better implementation method. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions for causing a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in various embodiments of the present application.

[0196] Obviously, the above-described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The accompanying drawings show the preferred embodiments of the present application, but do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosed content of the present application more thorough and comprehensive. Although the present application has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions described in the foregoing specific embodiments, or perform equivalent replacements for some of the technical features. Any equivalent structure directly or indirectly using the content of the specification and drawings of the present application in other related technical fields is similarly within the scope of the patent protection of the present application.

Claims

1. A data recommendation method based on artificial intelligence, characterized in that, It includes the following steps: Collect the input content and operation behavior data of the user in the interaction interface; Based on the preset recognition rules, conduct appeal recognition on the input content and the operation behavior data to obtain the user type of the user; Judge whether the user type is a car-purchasing user; If so, collect the vehicle category data and the user's own condition information input by the user; Obtain the historical relevant data and market dynamic information corresponding to the vehicle category data; Based on the preset recommendation model, conduct recommendation processing on the vehicle category data, the user's own condition information, the historical relevant data, and the market dynamic information to obtain the corresponding recommended vehicle model results; Return the recommended vehicle model results to the user.

2. The data recommendation method based on artificial intelligence according to claim 1, wherein Before the step of conducting recommendation processing on the vehicle category data, the user's own condition information, the historical relevant data, and the market dynamic information based on the preset recommendation model to obtain the corresponding recommended vehicle model results, it further includes: Obtain the initial vehicle model data from the preset data source; Construct sample data based on the initial vehicle model data; Invoke the pre-constructed initial large model; Based on the preset model training strategy, use the sample data to train the initial large model to obtain the corresponding first large model; Based on the preset evaluation index, conduct performance evaluation on the first large model to obtain the corresponding performance evaluation result; Based on the performance evaluation result, conduct model optimization processing on the first large model to obtain the optimized second large model; Use the second large model as the recommendation model.

3. The data recommendation method based on artificial intelligence according to claim 2, wherein Before the step of invoking the pre-constructed initial large model, it further includes: Obtain the preset distributed architecture and generate a specified model based on the distributed architecture; Set an input layer including a first preset number of input channels in the specified model to obtain the corresponding first generated model; Set a second preset number of hidden layers in the first generated model to obtain the corresponding second generated model; Obtain the preset application scenario; Set an output layer corresponding to the application scenario in the second generated model to obtain the corresponding third generated model; Use the third generated model as the initial large model.

4. The data recommendation method based on artificial intelligence according to claim 2, wherein, Before the step of using the sample data to train the initial large model based on the preset model training strategy to obtain the corresponding first large model, it further includes: Obtain the preset supervised learning training strategy, unsupervised learning training strategy, and reinforcement learning training strategy; Conduct combined processing on the supervised learning training strategy, the unsupervised learning training strategy, and the reinforcement learning training strategy to obtain the corresponding combined training strategy; Use the combined training strategy as the model training strategy.

5. The data recommendation method based on artificial intelligence according to claim 1, wherein The step of returning the recommended vehicle model results to the user specifically includes: Conduct scoring processing on all the recommended vehicle model results to obtain the corresponding scoring results; Based on the scoring results, conduct sorting processing on all the recommended vehicle model results to obtain the corresponding recommended vehicle model list; Conduct supplementary processing on the decision-making basis of the recommended vehicle model list to obtain the corresponding target recommended vehicle model list; Obtain the preset target feedback method; Based on the target feedback method, return the target recommended vehicle model list to the user.

6. The data recommendation method based on artificial intelligence according to claim 1, wherein After the step of determining whether the user type is a vehicle purchase user, it further includes: If the user type is not a vehicle purchase user, then determine whether the user type is a learning user; If so, obtain the consultation question type from the input content; Obtain learning resources corresponding to the consultation question type from a preset vehicle model knowledge base; Construct a learning path corresponding to the learning resources; Send the learning resources and the learning path to the user.

7. The data recommendation method based on artificial intelligence according to claim 1, characterized in that After the step of performing recommendation processing on the vehicle category data, the self-condition information, the historical related data, and the market dynamic information based on a preset recommendation model to obtain corresponding recommended vehicle model results, it further includes: Extract key index data from the recommended vehicle model results based on a preset index type; Invoke a preset chart drawing tool; Perform chart drawing on the key index data based on the chart drawing tool to obtain a corresponding first chart; Perform text explanation processing on the first chart based on the recommended vehicle model results to obtain a corresponding second chart; Perform display processing on the second chart.

8. An artificial intelligence-based data recommendation device, characterized in that, It includes: A first collection module, configured to collect the input content and operation behavior data of the user in the interaction interface; An identification module, configured to perform appeal identification on the input content and the operation behavior data based on a preset identification rule to obtain the user type of the user; A first judgment module, configured to judge whether the user type is a vehicle purchase user; A second collection module, configured to, if so, collect the vehicle category data and self-condition information input by the user; A first acquisition module, configured to acquire historical related data and market dynamic information corresponding to the vehicle category data; A recommendation module, configured to perform recommendation processing on the vehicle category data, the self-condition information, the historical related data, and the market dynamic information based on a preset recommendation model to obtain corresponding recommended vehicle model results; A return module, configured to return the recommended vehicle model results to the user.

9. A computer device, characterized in that, It includes a memory and a processor, and computer-readable instructions are stored in the memory. When the processor executes the computer-readable instructions, the steps of the data recommendation method based on artificial intelligence according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that, Computer-readable instructions are stored on the computer-readable storage medium. When the computer-readable instructions are executed by the processor, the steps of the data recommendation method based on artificial intelligence according to any one of claims 1 to 7 are implemented.

Images