Automobile brand recommendation method and system based on big data

Abstract

The invention discloses an automobile brand recommendation method and system based on big data, and relates to the technical field of big data, and the method comprises the steps: collecting automobile model parameters, selling prices and user scores under all automobile brand flags; determining scoring indexes according to the scores of the users, and deeply analyzing guide factors of the vehicle model parameters to the scoring indexes; establishing an automobile model recommendation model for the guide factors of the scoring indexes based on the automobile model parameters; using an automobile model recommendation model to recommend an automobile brand most conforming to the user's expectation, and listing recommended automobile models of which the selling prices under the brand flag are within the user budget; and continuously optimizing the automobile model recommendation model according to the attention of the user on the recommended automobile model. According to the invention, customized automobile brand selection is provided for the user based on the preference, demand and budget of the user, so that the time and energy of the user are saved; the vehicle purchase intention of a potential customer can be identified, and the user is guided to pay attention to the vehicle brand most meeting the demand, so that the sales conversion rate is improved.

Description

Technical Field

[0001] The present invention relates to the field of big data technology, and particularly to a method and system for recommending automobile brands based on big data. Background Art

[0002] As a general means of transportation, automobiles have become an essential part of life. Before purchasing an automobile, customers will do a lot of preparatory work to buy a new car suitable for themselves. For example, they will check the official websites of multiple automobile brands to understand the prices and parameters of automobiles, and look for the sensory and driving experiences of users who have bought them. However, some new brands or niche cars may be more suitable for customers and meet their expectations better, but they will not be retrieved by the public because of their low popularity, resulting in missed potential customers. Therefore, customers also miss a car that suits them. To solve the above problems, the present application provides a method and system for recommending automobile brands based on big data, which can accurately recommend automobile brands according to users' preferences. Summary of the Invention

[0003] The present invention provides a method for recommending automobile brands based on big data, including: Step1, collecting the model parameters, selling prices, and user ratings of each automobile brand; Step2, determining the corresponding rating indicators according to each user rating, and deeply analyzing the guiding factors of the model parameters for each rating indicator; Step3, establishing an automobile model recommendation model based on the guiding factors of the model parameters for each rating indicator; Step4, using the automobile model recommendation model to recommend the automobile brand that best meets the user's expectations, and listing the recommended models under this brand whose selling prices are within the user's budget; Step5, continuously optimizing the automobile model recommendation model according to the user's attention to the recommended models.

[0004] For the method for recommending automobile brands based on big data as described above, where the corresponding rating indicators are determined according to each user rating, and the guiding factors of the model parameters for each rating indicator are deeply analyzed, it is specifically divided into the following sub-steps: Processing each model parameter and the corresponding rating indicator into an analysis pair; Calculating the guiding factors of the model parameters for each rating indicator under each rating change according to the rating difference between the analysis pairs; Integrating the guiding factors of the model parameters for each rating indicator under each rating change to obtain a final value set.

[0005] For the method for recommending automobile brands based on big data as described above, where an automobile model recommendation model is established based on the guiding factors of the model parameters for each rating indicator, it is specifically divided into the following sub-steps: Collect the real consultation records and purchase records of users who have purchased cars under each brand, and process them into a training data set; Based on the guiding factors of each scoring index for vehicle models, create a vehicle model recommendation model, and use the training data set to train and optimize the created model; Introduce manual intervention to perform secondary optimization on the model to obtain the final vehicle model recommendation model.

[0006] A method for recommending automobile brands based on big data as described above, in which the vehicle model recommendation model is used to recommend the automobile brand that best meets the user's expectations, and list the recommended models under this brand whose prices are within the user's budget. Specifically, it is divided into the following sub-steps: The user inputs the expected values of various indicators of the car through the expected input window; The vehicle model recommendation model outputs the model that best meets the expected values according to the input expected values; Return the brand of the model output by the model as the recommended brand for this time, and retrieve the models under the recommended brand that are similar to the model output by the model and whose prices are within the user's budget and return them to the user.

[0007] The present invention also provides a vehicle brand recommendation system based on big data, including: a vehicle data acquisition module, a vehicle data analysis module, a vehicle model recommendation model establishment module, a recommended data generation module, and a feedback data application module; The vehicle data acquisition module is used to collect the vehicle model parameters, prices, and user scores under each vehicle brand; The vehicle data analysis module is used to determine the corresponding scoring indicators according to the scores of each user, and deeply analyze the guiding factors of vehicle model parameters for each scoring indicator; The vehicle model recommendation model establishment module is used to establish a vehicle model recommendation model based on the guiding factors of vehicle model parameters for each scoring indicator; The recommended data generation module is used to use the vehicle model recommendation model to recommend the vehicle brand that best meets the user's expectations, and list the recommended models under this brand whose prices are within the user's budget; The feedback data application module is used to continuously optimize the vehicle model recommendation model according to the user's attention to the recommended models.

[0008] The beneficial effects achieved by the present invention are as follows: It can provide users with customized vehicle brand selection based on the user's preferences, needs, and budget, thereby saving the user's time and energy; it can identify the purchase intention of potential customers and guide users to pay attention to the vehicle brands that best meet their needs, thereby improving the sales conversion rate; it helps the positioning and differentiation of each brand in the market and enhances the brand competitiveness. Description of the Drawings

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to these drawings.

[0010] Figure 1 It is a flowchart of a method for recommending car brands based on big data provided in the first embodiment of the present application; Figure 2 It is a schematic diagram of a system for recommending car brands based on big data provided in the second embodiment of the present application. Detailed implementation manners

[0011] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0012] Embodiment 1

[0013] As Figure 1 shown, the first embodiment of the present application provides a method for recommending car brands based on big data, including: Step S10: Collect the model parameters, selling prices, and user ratings of each car brand; Use web crawler technology to obtain the model parameters and selling prices of each brand from the official websites or flagship stores of each car brand, and then store them in a structured database in the form of a table. Provide a Dianping window, and users who have registered the car brand recommendation system can rate the various performance aspects of their existing models. The rating indicators include, but are not limited to, power performance, fuel economy, handling performance, comfort performance, safety performance, and durability. System administrators can further refine and adjust the rating indicators according to the concerns of system users to make the recommended vehicles more meet the expectations of users.

[0014] Step S20: Determine the corresponding rating indicators according to each user rating, and deeply analyze the guiding factors of the model parameters for each rating indicator; Model parameters are the main reference data for users when purchasing a car, and these data have a certain guiding effect on the ratings given by users who have already purchased a car. For example, a small engine displacement of a model will naturally lead to good fuel economy but poor power performance. By quantifying this guiding effect, a general recommended range can be determined according to the specific needs of users. The analysis and quantification of the guiding effect specifically include the following sub-steps: Step S21: Process each vehicle model parameter and its corresponding scoring index into an analysis pair; The parameters of each vehicle model and its corresponding scoring data will be combined into an analysis pair, expressed as: , where respectively represent the various vehicle model parameters of a vehicle model, respectively represent the mean values of the respective scoring indices of this vehicle model, n is the total number of vehicle model parameters, and m is the total number of scoring indices.

[0015] Step S22: Calculate the guiding factors of vehicle model parameters for each scoring index under each scoring change based on the scoring differences between the analysis pairs; Starting from the first analysis pair, successively substitute the data of each pair of adjacent analysis pairs into the guiding factor calculation formula: to obtain the guiding factors of vehicle model parameters for each scoring index under the current scoring change, and store them in the data set E. The calculation result set E for each time is recorded in real-time in a large data set C, where is the value of the i-th vehicle model parameter in the (d + 1)-th analysis pair, is the value of the i-th vehicle model parameter in the d-th analysis pair, i takes values from 1 to n, and n is the total number of vehicle model parameters, is the value of the j-th scoring index in the (d + 1)-th analysis pair, is the value of the j-th scoring index in the d-th analysis pair, j takes values from 1 to m, and m is the total number of scoring indices, is the identifier indicating whether the i-th vehicle model parameter has a guiding effect on the j-th scoring index. If so, it is 1; otherwise, it is 0.

[0016] Step S23: Integrate the guiding factors of vehicle model parameters for each scoring index under each scoring change to obtain a final value set; Input the data in the large data set C into the integration function F(C), and the final guiding factor set of vehicle model parameters for each scoring index can be returned. Among them, the integration function F(C) is expressed as: where represents the value of the j-th scoring index in the v-th subset of the large data C, represents the maximum value of the j-th scoring index in the large data C, represents the minimum value of the j-th scoring index in the large data set C, v takes values from 1 to q, and q is the total number of subsets in the large data C, j takes values from 1 to m, and m is the total number of scoring indices, is used to control the integration weight of the average quantity, is used to control the integration weight of the median quantity. The return value of the integration function F(C) is expressed as , where is the final guiding factor of vehicle model parameters for each scoring index.

[0017] Step S30: Establish an automotive model recommendation model based on the guiding factors of each scoring index for vehicle models; Based on the guiding factors of each scoring index for vehicle models, a general recommendation range can be determined (this can make up for the problem of insufficient training data and inaccurate models caused by too little purchase data). Then, combined with the transaction data provided by each brand, an automotive model recommendation model is trained to further narrow this recommendation range. Specifically: Step S31: Collect the real consultation records and purchase records of users who have purchased vehicles under each brand, and process them into a training data set; The consultation records contain the vehicle purchase expectation information of users. Use natural language processing technology to extract this information and quantify it based on the previously preset scoring indexes. The following is an optional quantification process: Associate some common expression terms with the scoring indexes. For example, the expressions for power performance can include: fast acceleration, strong climbing ability, small wind resistance coefficient, etc.; then use a similarity algorithm to calculate the similarity between the expectation information and the associated terms of each scoring index. If the similarity is higher than the threshold, it is determined that the user needs this performance. Then, according to how much the similarity exceeds the threshold and the frequency of occurrence, determine the user's expectation value for this index. The expectation value and the scoring value (the value of the scoring index) can be the same or there can be a certain conversion rate. In this embodiment, they are the same. Only such a quantification method can make full use of the previously obtained guiding factors when establishing the model.

[0018] After obtaining the quantification result of the consultation record, combine it with the corresponding vehicle model in the purchase record to form an input-output pair, expressed as where is the input, that is, the user's expectation value for each scoring index, and k is the output, that is, the subscript of the vehicle model finally purchased by the user. Finally, organize all input-output pairs into the training data set.

[0019] Step S32: Create an automotive model recommendation model based on the guiding factors of each scoring index for vehicle models, and use the training data set to train and optimize the created model; The mathematical expression of the automotive model recommendation model is: where is the recommendation coefficient of vehicle model k, is the identifier indicating whether the i-th vehicle model parameter has a guiding effect on the j-th scoring index. If it does, it is 1; otherwise, it is 0. is the i-th vehicle model parameter of the k-th vehicle model, is the expectation value of the user to be recommended for the j-th scoring index, is the guiding factor of the j-th scoring index, Used to return the value of k when the calculation result of the expression in parentheses is the smallest. The value of ki ranges from 1 to kn, where kn is the number of vehicle model parameters disclosed for vehicle model k. The value of j ranges from 1 to m, where m is the total number of scoring metrics. The value of k ranges from 1 to w, where w is the total number of vehicle models. The value of k is updated after each calculation is completed, and Y is the output of the model.

[0020] Use the training dataset to train the created vehicle model recommendation model. During the training process, continuously adjust the parameters in the model until the difference between the model output result and the actual result reaches the minimum.

[0021] Step S33: Introduce manual intervention to perform secondary tuning on the model to obtain the final vehicle model recommendation model; Prepare an expected input window to display each expected metric (consistent with the scoring metric), and then a group of testers fill in the expected values for each expected metric through this window. The vehicle model recommendation model outputs the corresponding recommended vehicle models according to the input expected values. The testers verify whether the output recommended vehicle models meet the expectations. If more than 90% of the outputs meet the requirements, the verification passes; otherwise, increase the training weights of the data that fails to pass the verification in the training dataset. After the model is retrained, perform verification and tuning again until the verification passes.

[0022] Step S40: Use the vehicle model recommendation model to recommend the vehicle brand that best meets the user's expectations, and list the recommended vehicle models under this brand whose prices are within the user's budget; Since the vehicle model recommendation model can now recommend the vehicle model that best meets the user's expectations, the brand to which this vehicle model belongs is regarded as the vehicle brand that best meets the user's expectations. The specific output process can be divided into the following sub-steps: Step S41: The user inputs their expected values for each vehicle metric through the expected input window; The metrics here are consistent with the preset scoring metrics in Step S10, but the obtained values will be input as expected values to the vehicle model recommendation model. The expected input window also includes a budget input box for obtaining the user's car purchase budget.

[0023] Step S42: The vehicle model recommendation model outputs the vehicle model that best meets the expected values according to the input expected values; Step S43: Return the brand of the vehicle model output by the model as the recommended brand for this time, and retrieve the vehicle models under the recommended brand that are similar to the vehicle model output by the model and whose prices are within the user's budget and return them to the user; Take the vehicle model output by the model as the main promoted vehicle and place it in the first row of the recommendation list. Then, arrange them in descending order of similarity to the main promoted vehicle below. The length of the recommendation list is set according to requirements. Only the basic information of the vehicle model, such as pictures, selling prices, release dates, vehicle models, etc., is displayed in each row of the recommended data. Specific information needs to be viewed by clicking into the details page, where all the publicly available vehicle model parameters of the vehicle model are displayed.

[0024] Step S50: Continuously optimize the vehicle model recommendation model according to the user's attention to the recommended vehicle models; After the user receives the recommended vehicle brands and models, they need to click into the details page to see the specific vehicle model data. Then, the user's attention to the recommended vehicle models can be reflected from the click-through rate and browsing duration. Regularly collect the user's input data and attention data, and then use these data to optimize the vehicle model recommendation model, so that the recommended results output are more in line with the user's concerns, thus achieving further optimization of the model.

[0025] Embodiment 2

[0026] As Figure 2 shown, Embodiment 2 of the present application provides a big data-based vehicle brand recommendation system, including: a vehicle data acquisition module 21, a vehicle data analysis module 22, a vehicle model recommendation model establishment module 23, a recommended data generation module 24, and a feedback data application module 25; The vehicle data acquisition module 21 is used to collect the vehicle model parameters, selling prices, and user ratings under each vehicle brand; Use web crawler technology to obtain the vehicle model parameters and selling prices under each brand from the official websites or flagship stores of each vehicle brand, and then store them in a structured database in the form of a table. Provide a Dianping window, and users who have registered the vehicle brand recommendation system can rate the various performance aspects of their existing vehicle models. The rating indicators include but are not limited to power performance, fuel consumption economy, handling performance, comfort performance, safety performance, and durability. System administrators can further refine and adjust the rating indicators according to the concerns of system users, so that the recommended vehicles better meet the user's expectations.

[0027] The vehicle data analysis module 22 is used to determine the corresponding rating indicators according to each user rating and deeply analyze the guiding factors of vehicle model parameters for each rating indicator; specifically including: an analysis pair generation sub-module, a guiding factor calculation sub-module, and a guiding factor integration sub-module; 1. The analysis pair generation sub-module is used to process each vehicle model parameter and the corresponding rating indicator into an analysis pair; The parameters of each vehicle model and the corresponding rating data will be combined into an analysis pair, expressed as: , where respectively represent the various vehicle model parameters of a vehicle model, respectively represent the mean values of each scoring index of the vehicle model, n is the total number of vehicle model parameters, and m is the total number of scoring indexes.

[0028] 2. The guiding factor calculation sub-module is used to calculate the guiding factors of vehicle model parameters for each scoring index under each scoring change according to the scoring differences between the analysis pairs; Starting from the first analysis pair, the data of each pair of adjacent analysis pairs are successively substituted into the guiding factor calculation formula: to obtain the guiding factors of vehicle model parameters for each scoring index under the current scoring change, and store them in the data set E. The calculation result set E of each time is recorded in a large data set C in real time, where is the value of the i-th vehicle model parameter in the (d + 1)-th analysis pair, is the value of the i-th vehicle model parameter in the d-th analysis pair, i takes values from 1 to n, and n is the total number of vehicle model parameters, is the value of the j-th scoring index in the (d + 1)-th analysis pair, is the value of the j-th scoring index in the d-th analysis pair, j takes values from 1 to m, and m is the total number of scoring indexes, is the identifier indicating whether the i-th vehicle model parameter has a guiding effect on the j-th scoring index. If it is, it is 1; otherwise, it is 0.

[0029] 3. The guiding factor integration sub-module is used to integrate the guiding factors of vehicle model parameters for each scoring index under each scoring change to obtain a final value set; By inputting the data in the large data set C into the integration function F(C), the final guiding factor set of vehicle model parameters for each scoring index can be returned, where the integration function F(C) is expressed as: where represents the value of the j-th scoring index in the v-th subset of the large data C, represents the maximum value of the j-th scoring index in the large data C, represents the minimum value of the j-th scoring index in the large data set C. v takes values from 1 to q, and q is the total number of subsets in the large data C. j takes values from 1 to m, and m is the total number of scoring indexes, is used to control the integration weight of the average quantity, is used to control the integration weight of the median quantity. The return value of the integration function F(C) is expressed as where is the final guiding factor of vehicle model parameters for each scoring index.

[0030] The vehicle model recommendation model establishment module 23 is used to establish a vehicle model recommendation model based on the guiding factors of vehicle model parameters for each scoring index; specifically including: a training data set creation sub-module, a model training sub-module, and a model verification sub-module; 1. Training dataset creation sub-module, which is used to collect the real consultation records and purchase records of users who have bought cars under each brand, and process them into a training dataset; The consultation records contain the car purchase expectation information of users. Use natural language processing technology to extract this information and quantify it based on the previously preset scoring metrics. The following is an optional quantification process: Associate some common expression terms with the scoring metrics. For example, the expressions for power performance can include: fast acceleration, strong climbing ability, small drag coefficient, etc.; then use a similarity algorithm to calculate the similarity between the expectation information and the terms associated with each scoring metric. If the similarity is higher than the threshold, it is determined that the user needs this performance. Then, determine the user's expectation value for this metric according to how much the similarity exceeds the threshold and the frequency of occurrence. The expectation value and the score value (the value of the scoring metric) can be the same or there can be a certain conversion rate. In this embodiment, they are the same. Only such a quantification method can make full use of the previously obtained guiding factors when building the model.

[0031] After obtaining the quantification result of the consultation record, combine it with the corresponding vehicle model in the purchase record to form an input-output pair, expressed as , where is the input, that is, the user's expectation value for each scoring metric, k is the output, that is, the subscript of the vehicle model finally purchased by the user. Finally, organize all the input-output pairs into the training dataset.

[0032] 2. Model training sub-module, which is used to train and optimize the created model using the training dataset; The mathematical expression of the vehicle model recommendation model is: , where is the recommendation coefficient of vehicle model k, is the identifier indicating whether the i-th vehicle model parameter of the j-th scoring metric has a guiding effect. If it does, it is 1, otherwise it is 0, is the i-th vehicle model parameter of the k-th vehicle model, is the expectation value of the user to be recommended for the j-th scoring metric, is the guiding factor of the j-th scoring metric, is used to return the k value when the calculation result of the expression in the parentheses is the smallest. ki takes values from 1 to kn, kn is the number of vehicle model parameters publicly disclosed for vehicle model k, j takes values from 1 to m, m is the total number of scoring metrics, k takes values from 1 to w, w is the total number of vehicle models, and the k value is updated after each calculation is completed. Y is the output of the model.

[0033] Use the training dataset to train the created vehicle model recommendation model, and continuously adjust the parameters in the model during the training process until the difference between the model output result and the actual result reaches the minimum.

[0034] 3. The model verification sub-module is used to introduce manual intervention to perform secondary optimization on the model to obtain the final car model recommendation model; Prepare an expected input window to display each expected index (consistent with the scoring index), and then a group of testers fill in the expected values for each expected index through this window. The car model recommendation model outputs the corresponding recommended models according to the input expected values. The testers verify whether the output recommended models meet the expectations. If more than 90% of the outputs meet the requirements, the verification passes; otherwise, the training weights of the data that fails the verification in the training dataset will be increased. After the model is retrained, verification and optimization are performed again until the verification passes.

[0035] The recommendation data generation module 24 is used to use the car model recommendation model to recommend the car brand that best meets the user's expectations, and list the recommended models under this brand whose prices are within the user's budget; Since the car model recommendation model can now recommend the model that best meets the user's expectations, the brand to which this model belongs is regarded as the car brand that best meets the user's expectations. The specific output process can be divided into the following sub-steps: I. The user inputs their expected values for various car indicators through the expected input window; The indicators here are consistent with the preset scoring indicators in step S10, but the obtained values will be used as expected values and input into the car model recommendation model. The expected input window also includes a budget input box for obtaining the user's car purchase budget.

[0036] II. The car model recommendation model outputs the model that best meets the expected values according to the input expected values; III. Return the brand of the model output by the model as the recommended brand for this time, and retrieve the models under the recommended brand that are similar to the model output by the model and whose prices are within the user's budget and return them to the user; Take the model output model as the main promoted model and place it in the first row of the recommendation list. Below, arrange them in descending order of similarity to the main promoted model in turn. The length of the recommendation list is set as needed. Each row of recommended data only displays the basic information of the model, such as pictures, prices, release dates, models, etc. The specific information needs to be clicked to enter the details page to view, and all publicly available model parameters of the model are displayed on the details page.

[0037] The feedback data application module 25 is used to continuously optimize the car model recommendation model according to the user's attention to the recommended models; After receiving the recommended car brands and models, the user needs to click into the details page to view the specific model data. Then, the user's attention to the recommended models can be reflected from the click-through rate and browsing duration. Regularly collect the user's input data and attention data, and then use these data to optimize the car model recommendation model, so that the recommended results output are more in line with the user's focus points, thereby achieving further optimization of the model.

[0038] Corresponding to the above embodiment, an embodiment of the present invention provides a computer storage medium, including: at least one memory and at least one processor; The memory is used to store one or more program instructions; The processor is used to run one or more program instructions to execute a method for recommending car brands based on big data.

[0039] Corresponding to the above embodiment, an embodiment of the present invention provides a computer-readable storage medium. The computer storage medium contains one or more program instructions, and the one or more program instructions are used to be executed by the processor to execute a method for recommending car brands based on big data.

[0040] The embodiment disclosed by the present invention provides a computer-readable storage medium. Computer program instructions are stored in the computer-readable storage medium. When the computer program instructions run on a computer, the computer is enabled to execute the above-mentioned method for recommending car brands based on big data.

[0041] In the embodiment of the present invention, the processor may be an integrated circuit chip with signal processing capabilities. The processor may be a general-purpose processor, a digital signal processor (DSP for short), an application-specific integrated circuit (ASIC for short), a field-programmable gate array (FPGA for short), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0042] The various methods, steps, and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general-purpose processor can be a microprocessor, or the processor can also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present invention can be directly embodied as being executed and completed by a hardware decoding processor, or can be executed and completed by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The processor reads the information in the storage medium and combines its hardware to complete the steps of the above method.

[0043] The storage medium can be a memory, for example, it can be a volatile memory or a non-volatile memory, or can include both a volatile and a non-volatile memory.

[0044] Among them, the non-volatile memory can be a read-only memory (ROM for short), a programmable read-only memory (PROM for short), an erasable programmable read-only memory (EPROM for short), an electrically erasable programmable read-only memory (EEPROM for short), or a flash memory.

[0045] The volatile memory can be a random access memory (RAM for short), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM for short), dynamic random access memory (DRAM for short), synchronous dynamic random access memory (SDRAM for short), double data rate synchronous dynamic random access memory (DDR SDRAM for short), enhanced synchronous dynamic random access memory (ESDRAM for short), synchronous link dynamic random access memory (SLDRAM for short), and direct rambus random access memory (DRRAM for short).

[0046] The storage medium described in the embodiments of the present invention is intended to include but not be limited to these and any other suitable types of memories.

[0047] Those skilled in the art should be able to realize that in one or more of the above examples, the functions described in the present invention can be implemented by a combination of hardware and software. When applying software, the corresponding functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. The computer-readable medium includes computer storage media and communication media, where the communication media includes any medium that facilitates the transmission of a computer program from one place to another. The storage media can be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0048] The specific embodiments described above further elaborate on the purpose, technical solutions, and beneficial effects of the present invention. It should be understood that the above description is only for the specific embodiments of the present invention and is not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made on the basis of the technical solutions of the present invention should be included in the protection scope of the present invention.

Claims

1. A car brand recommendation method based on big data, characterized in that: include: Step 1. Collect the model parameters, prices, and user ratings of each car brand; Step 2: Determine the corresponding scoring indicators based on the scores of each user, and deeply analyze the guiding factors of vehicle model parameters on each scoring indicator; Step 3: Establish a car model recommendation model based on the guiding factors of each scoring index based on the model parameters; Step 4: Use the car model recommendation model to recommend the car brand that best meets the user's expectations, and list the recommended models of the brand that are within the user's budget; Step 5. Continue to optimize the car model recommendation model based on users' attention to recommended models.

2. The automobile brand recommendation method based on big data according to claim 1, characterized in that: Determine the corresponding scoring indicators based on the scores of each user, and deeply analyze the guiding factors of vehicle model parameters on each scoring indicator, which is specifically divided into the following sub-steps: Treat each vehicle model parameter and the corresponding scoring index as an analysis pair; Based on the score difference between the analysis pairs, the guiding factors of the vehicle model parameters on each score indicator under each score change are calculated; The guiding factors of vehicle model parameters on each scoring index under each scoring change are integrated to obtain a final value set.

3. The automobile brand recommendation method based on big data according to claim 1, characterized in that: The car model recommendation model is established based on the guiding factors of each scoring index based on the model parameters, which is specifically divided into the following sub-steps: Collect the real consultation records and purchase records of car-buying users of various brands and process them into training data sets; Create a car model recommendation model based on the guiding factors of each scoring index based on the model parameters, and use the training data set to train and tune the created model; Human intervention is introduced to perform secondary tuning on the model to obtain the final car model recommendation model.

4. The automobile brand recommendation method based on big data according to claim 1, characterized in that: Use the car model recommendation model to recommend the car brand that best meets the user's expectations, and list the recommended models of the brand that are within the user's budget. The specific steps are as follows: The user enters his or her expected values ​​for various indicators of the car through the expected input window; The car model recommendation model outputs the car model that best meets the expected value based on the input expected value; The brand of the model output vehicle model is returned as the recommended brand this time, and the models of the recommended brand that are similar to the model output vehicle model and whose prices are within the user's budget are retrieved and returned to the user.

5. A car brand recommendation system based on big data, characterized in that: include: Automobile data acquisition module, automobile data analysis module, automobile model recommendation model building module, recommendation data generation module, feedback data application module; The car data acquisition module is used to collect the model parameters, prices, and user ratings of each car brand; The car data analysis module is used to determine the corresponding scoring indicators according to the scores of each user, and deeply analyze the guiding factors of the vehicle model parameters on each scoring indicator; A car model recommendation model building module is used to build a car model recommendation model based on the guiding factors of each scoring index based on the car model parameters; The recommendation data generation module is used to use the car model recommendation model to recommend the car brand that best meets the user's expectations, and list the recommended models of the brand that are within the user's budget; The feedback data application module is used to continuously optimize the automobile model recommendation model based on the user's attention to the recommended models.

6. The automobile brand recommendation system based on big data according to claim 5, characterized in that: The automobile data analysis module specifically includes: an analysis pair generation submodule, a guide factor calculation submodule, and a guide factor integration submodule; An analysis pair generation submodule is used to process each vehicle model parameter and the corresponding scoring index into an analysis pair; The guiding factor calculation submodule is used to calculate the guiding factors of the vehicle model parameters for each scoring index at each score change according to the score difference between the analysis pairs; The guiding factor integration submodule is used to integrate the guiding factors of vehicle model parameters for each scoring index under each score change to obtain a final value set.

7. The automobile brand recommendation system based on big data according to claim 5, characterized in that: The automobile model recommendation model building module specifically includes: training data set creation submodule, model training submodule, and model verification submodule; The training data set creation submodule is used to collect the real consultation records and purchase records of car-buying users of various brands and process them into training data sets; The model training submodule is used to train and tune the created model using the training dataset; The model verification submodule is used to introduce manual intervention to perform secondary tuning on the model to obtain the final car model recommendation model.

8. A computer storage medium, characterized in that: include: at least one memory and at least one processor; A memory for storing one or more program instructions; A processor, used to run one or more program instructions to execute a car brand recommendation method based on big data as described in any one of claims 1 to 4.

Images