A method for recommending residential car advertising locations based on Internet of Vehicles data
Through the advertising site selection recommendation method based on Internet of Vehicles data, combined with the community purchasing potential, infrastructure capabilities and community matching, the advertising recommendation index is calculated, which solves the problem of low advertising delivery efficiency in the existing technology and realizes the new energy supply The precise reach of car potential customers and the improvement of advertising.
Patent Information
- Application Number
- CN202311754398.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-12-20
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2043-12-20
AI Technical Summary
The existing methods for recommending automotive advertising site selection lack a complete system based on Internet of Vehicles data, making it difficult to accurately reach new energy vehicle potential customers, resulting in low advertising delivery efficiency.
By obtaining information on the purchasing potential of the community, infrastructure capacity configuration degree and community matching degree, the advertising recommendation index is calculated using a pre-trained learning model to achieve targeted advertising delivery.
It has achieved accurate analysis and advertising delivery of high-rated communities, enhanced the value of advertising clues, helped car companies to tap into user groups with high potential customers, and improved the accuracy of advertising delivery.
Smart Images

Figure CN117876041B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of new energy automobile marketing, and in particular relates to a method for recommending residential automobile advertising sites based on Internet of Vehicles data. Background Art
[0002] As the new energy vehicle industry continues to mature, the sales of new energy vehicles have gradually increased, and the sales volume of new energy vehicles has exceeded 20% of the total vehicle sales. New energy vehicle products are facing severe homogeneous competition, and the marketing expenses invested by the OEMs are also high. Compared with store site selection, the site selection range is smaller and the focus is mainly on customer sources. The site selection range for community advertising is larger, and regional users have the characteristics of convergent preferences. It is necessary to consider the purchasing customer volume and customer matching, and reach community users in a targeted manner.
[0003] The existing methods for recommending the location of automobile advertisements are mainly to place advertisements in the community and the charging conditions of the community advertisement space by locating the community population and matching the configuration of the automobile. For example, the price range of the community residents' automobile purchase is estimated to match the housing price of the community. At present, the location selection of residential areas is mainly based on traditional theoretical basis combined with partial data, and a complete set of index system for the location selection of new energy automobile residential areas has not yet been formed. Summary of the invention
[0004] In view of this, the purpose of the present invention is to propose a method for recommending residential automobile advertising sites based on Internet of Vehicles data, so as to achieve targeted advertising placement in high-scoring communities, help automobile companies to explore communities with high potential customers and user matching for advertising placement in home elevators, etc., improve the value of advertising leads, and assist automobile companies in accurately reaching target users.
[0005] In order to achieve the above object, the present invention provides the following technical solutions:
[0006] The present invention provides a method for recommending a residential automobile advertisement site based on Internet of Vehicles data, comprising:
[0007] Get the first category information of multiple residential communities in a specified city;
[0008] Based on the first type of information, multiple first type indicators are determined, and after the first type indicators are standardized, the coefficient of variation of the single indicator is used to calculate the community purchase potential index and the infrastructure capacity configuration index;
[0009] Determine a plurality of second-category indicators based on the first-category information, input a pre-trained learning model into the second-category indicators and output a community matching index;
[0010] According to the community purchasing potential index, infrastructure capacity configuration index and community community matching index, the advertising recommendation index of the corresponding community is calculated based on the preset weight coefficient;
[0011] Determine the advertising recommendation index corresponding to each of all residential communities in the designated city, and select recommended addresses in order of the advertising recommendation index.
[0012] Preferably, the first type of information includes community purchase potential information, community infrastructure capacity information and community community information, wherein:
[0013] The purchase potential information of the community includes the number of households in the community, the average price of vehicles in the community, the total number of new energy vehicles, the number of new energy vehicles in the same market segment, the average age of new energy vehicles purchased, the plot ratio of the community, the greening rate of the community, the distance between the community and the subway station, the number of households in the community, the number of buildings in the community, and the number of parking spaces in the community;
[0014] The infrastructure capacity information of the community includes the proportion of private charging piles, the distance to the nearest public charging pile, the proportion of fast charging piles in the area, the car-to-pile ratio, the convenience of charging, and the waiting time for charging;
[0015] Community information includes gender, age, whether there are children, average driving time per trip, average driving mileage per trip, average daily mileage, charging start SOC, endurance / commuting distance, proportion of fast charging times, and number of full-charge endurance times.
[0016] Preferably, after the first type of indicators are standardized, the community purchase potential index and the infrastructure capacity configuration index are calculated based on the coefficient of variation of the single indicator, respectively, including:
[0017] The first category of indicators is subdivided into the third category of indicators and the fourth category of indicators according to the preset indicator classification standards;
[0018] Based on the third type of indicators, the purchase potential index of the community is calculated by the first type of calculation method. Based on the fourth type of indicators, the infrastructure capacity configuration index is calculated by the first type of calculation method.
[0019] Among them, the community purchase potential index calculated based on the third type of indicators through the first type of calculation method includes:
[0020] The range normalization method is used to find the maximum and minimum values of a single indicator in the third category of indicators, and the indicators are normalized to the maximum and minimum values:
[0021]
[0022] in, is the corresponding j-th indicator after standardization. is the jth index before standardization. represents the largest j-th indicator determined among all the communities in the specified city, It represents the smallest j-th indicator determined among all the communities in the specified city;
[0023] Calculate the coefficient of variation for a single indicator:
[0024]
[0025] in, represents the coefficient of variation of the j-th indicator, represents the standard deviation of the jth indicator, represents the average value of the jth indicator;
[0026] Calculate the community purchasing potential index:
[0027]
[0028] in, Indicates the purchasing potential index of the community. represents the correction coefficient of the jth indicator;
[0029] The infrastructure capacity configuration index calculated based on the fourth type of indicators using the first type of calculation method includes:
[0030] The range normalization method is used to find the maximum and minimum values of a single indicator in the fourth category of indicators, and the indicators are normalized to the maximum and minimum values:
[0031]
[0032] in, is the corresponding i-th indicator after standardization, is the corresponding i-th indicator before standardization, represents the largest i-th indicator determined among all the communities in a specified city, It represents the smallest i-th indicator determined among all the communities in the specified city;
[0033] Calculate the coefficient of variation for a single indicator:
[0034]
[0035] in, represents the coefficient of variation of the i-th indicator, represents the standard deviation of the i-th indicator, represents the average value of the i-th indicator;
[0036] Calculate the infrastructure capacity configuration index:
[0037]
[0038] in, represents the infrastructure capacity configuration index, Represents the correction coefficient of the i-th indicator.
[0039] Preferably, determining a plurality of second-category indicators based on the first-category information, inputting a pre-trained learning model into the second-category indicators and outputting a community matching index includes:
[0040] The second category of indicators is subdivided into the fifth category of indicators and the sixth category of indicators, and the fifth category of indicators is used as the independent variable and the sixth category of indicators is used as the dependent variable;
[0041] The feature factors are screened by random forest, and the importance of features is ranked according to the contribution of different features to the data set, and the indicator factors with higher importance are selected;
[0042] According to different cities, 70% of the data are selected as the experimental set and 30% as the test set;
[0043] Different regression algorithms are used to train the target value, and the model results are evaluated by mean square error, root mean square error, and mean absolute error. The underfitting and overfitting models are optimized after adjusting the parameters.
[0044] Select the optimal model, put the corresponding market segmentation data indicators into the model, and output the community matching index.
[0045] Preferably, according to the community purchasing potential index, the infrastructure capacity configuration index and the community community matching index, the advertising recommendation index of the corresponding community is calculated based on the preset weight coefficient, including:
[0046] Obtain the corresponding weight coefficients of the preset community purchase potential index, infrastructure capacity configuration index and community community matching index;
[0047] The advertising recommendation index of the community is calculated by the following formula:
[0048]
[0049] in, Indicates the purchase potential index of the community. represents the infrastructure capacity configuration index, represents the community matching index, , , They represent the weight coefficients of the community purchasing potential index, infrastructure capacity configuration index and community community matching index respectively. It represents the minimum value of the purchase potential index of all communities in each city. It represents the maximum value of the purchase potential index of all communities in each city. It represents the minimum value of the infrastructure capacity configuration index of all communities in each city. It represents the maximum value of the infrastructure capacity configuration index of all communities in each city. It represents the minimum value of the community matching index of all communities in each city. It represents the maximum value of the community matching index of all communities in each city.
[0050] The present invention has achieved at least the following beneficial effects:
[0051] 1. Analyze and explore the purchasing potential of communities, facilitate targeted advertising in high-scoring communities, help automakers explore communities with high potential customers and user matching to place advertisements in entrance elevators, etc., increase the value of advertising leads, and assist automakers in accurately reaching target users.
[0052] Other advantages, objectives and features of the present invention will be described in the following description and will be apparent to those skilled in the art to some extent, or those skilled in the art may be taught from the practice of the present invention. The objectives and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS
[0053] In order to make the purpose, technical solution and beneficial effects of the present invention clearer, the present invention provides the following drawings for illustration:
[0054] Figure 1 The present invention is a flowchart of the steps of a method for recommending a location for automobile advertisements in a residential area based on Internet of Vehicles data in an embodiment of the present invention. DETAILED DESCRIPTION
[0055] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.
[0056] The present invention provides a method for recommending a residential automobile advertisement site based on Internet of Vehicles data, referring to Figure 1 ,include:
[0057] Step S1, obtaining first category information of multiple residential communities in a designated city;
[0058] Step S2: determining a plurality of first-category indicators based on the first-category information, performing standardization on the first-category indicators, and calculating the community purchase potential index and the infrastructure capacity configuration index based on the coefficient of variation of the individual indicators;
[0059] Step S3: determining a plurality of second-category indicators based on the first-category information, inputting a pre-trained learning model into the second-category indicators and outputting a community matching index;
[0060] Step S4, calculating the advertising recommendation index of the corresponding community based on the preset weight coefficient according to the community purchase potential index, the infrastructure capacity configuration index and the community community matching index;
[0061] Step S5: determine the advertisement recommendation index corresponding to each of the residential communities in the designated city, and select the recommended address in the order of the advertisement recommendation index.
[0062] The working principle and beneficial effects of the above technical solution are as follows: starting from users who have purchased new energy vehicles, combined with the purchase potential of the community (the number of users willing to purchase new energy vehicles), infrastructure capacity configuration (the ability and convenience of vehicle charging), and community community matching requirements (configuration requirements for purchasing new energy vehicles), the dimensional data in different cities are integrated and modeled, and the community advertising placement index in different cities is comprehensively considered. Targeted advertising is placed in high-scoring communities, helping car companies to explore advertising placements such as entrance elevators in communities with high potential customers and user matching, thereby increasing the value of advertising leads and assisting car companies in accurately reaching target users.
[0063] In a preferred embodiment, the first type of information includes community purchase potential information, community infrastructure capacity information, and community community information, wherein:
[0064] The purchase potential information of the community includes the number of households in the community, the average price of vehicles in the community, the total number of new energy vehicles, the number of new energy vehicles in the same market segment, the average age of new energy vehicles purchased, the plot ratio of the community, the greening rate of the community, the distance between the community and the subway station, the number of households in the community, the number of buildings in the community, and the number of parking spaces in the community;
[0065] The infrastructure capacity information of the community includes the proportion of private charging piles, the distance to the nearest public charging pile, the proportion of fast charging piles in the area, the car-to-pile ratio, the convenience of charging, and the waiting time for charging;
[0066] Community information includes gender, age, whether there are children, average driving time per trip, average driving mileage per trip, average daily mileage, charging start SOC, endurance / commuting distance, proportion of fast charging times, and number of full-charge endurance times.
[0067] The working principle and beneficial effects of the above technical solution are as follows: through the community purchase potential information, the number of users with purchasing intention in the community and whether the overall purchasing power of the community is consistent with the vehicle are evaluated; through the community infrastructure capacity information, it is judged whether the customer's charging and private charging pile construction needs are met; through the community community information, it is evaluated whether the community users have similar preferences and whether the vehicle configuration matches the needs of the community users. The purchase potential of the community includes: the number of households in the community, the average price of the community, the total number of new energy vehicles, the number of new energy vehicles in the same market segment, the average age of new energy vehicles purchased, the community plot ratio, the community greening rate, and the distance between the community and the subway Station distance, number of households in the community, number of buildings in the community, number of parking spaces in the community, etc.; Community community matching degree: gender, age, whether there are children, average driving time per trip, average driving mileage per trip, average daily mileage, charging start SOC, endurance / commuting distance, proportion of fast charging times, number of full-charge endurance times, etc.; Infrastructure capacity configuration degree: proportion of private charging piles, distance to the nearest public charging pile, proportion of fast charging piles in the area, vehicle-to-pile ratio (number of charging piles in the area / total number of vehicles in the area), charging convenience (number of idle charging piles in the area / number of vehicles with charging needs in the area), charging waiting time (the time interval between arrival time and actual charging time), etc.
[0068] In a preferred embodiment, it also includes data extraction of community purchase potential information, community infrastructure capacity information, and community community information, clarifies the data calculation caliber, and extracts community data, vehicle data, charging station data, and population data respectively. Extract the original data of the offline map, select the static information of the POI points of the residential community as the map points; extract the driving and charging clips of new energy vehicles in the community for the past month; extract the basic static data of the charging station; extract the natural and social attributes of the population and user habits. The data is based on the community as the dimension, and the data caliber of the community purchase potential is as follows:
[0069] Number of households in the community: select the number of households field of the community data;
[0070] Average price of community: select the average monthly price of the community in the community data;
[0071] Total number of new energy vehicles: select the total number of new energy vehicles that are permanently stationed in the community in that month (permanent judgment: the community with the most nighttime stays, if the number of stays is the same, select the community with the longest stay time);
[0072] Number of new energy vehicles in the same market segment: select the number of permanent new energy vehicles in the market segment in the current month (the market segment is divided and combined by power type and price range. The power type is divided into BEV and PHEV, and the price range is divided into less than 100,000, 100,000-200,000, 200,000-300,000, 300,000-500,000, and more than 500,000);
[0073] Average age of new energy vehicles purchased: Calculate the average age of new energy vehicles purchased in the community, retaining one decimal place;
[0074] Community Floor Area Ratio: Select the Floor Area Ratio field in the community data;
[0075] Community greening rate: select the greening rate field in the community data;
[0076] Distance between the community and the subway station: select the distance to the nearest subway station in the community data; Number of households in the community: select the number of households field in the community data;
[0077] Number of buildings in the community: select the number of buildings in the community data;
[0078] Number of parking spaces in the community: select the number of parking spaces field in the community data;
[0079] The data caliber of infrastructure capacity configuration is as follows:
[0080] The proportion of private charging piles: the number of new energy vehicles with private charging piles in the community / the total number of new energy vehicles in the community;
[0081] Distance to the nearest public pole: The distance between the community and the nearest public pole, unit: meter
[0082] The proportion of fast charging piles in the region: the number of fast charging piles in the region / the number of charging piles in the region
[0083] (Region determination: Traditional region determination is a fixed polygon (h3 hexagon or square), and the consistent area is convenient for horizontal comparison, but the community may be divided into different polygons. This patent adopts the concept of expanding the community outward by 800 meters. The choice of 800 meters is based on the fact that it is more acceptable for residents to walk 10 minutes and about 800 meters to the charging pile for charging)
[0084] Car-to-charging pile ratio: number of charging piles in a region / total number of vehicles in a region
[0085] Charging demand satisfaction - indicator 1 (charging convenience): the number of idle charging piles in the area / the number of vehicles with charging needs in the area
[0086] Charging demand satisfaction - indicator 2 (charging waiting time): the interval between the time when the resident vehicles in the community arrive at the charging station and the time when the vehicles actually start charging, unit: minutes;
[0087] The data caliber of the community matching degree is as follows:
[0088] Gender: the percentage of male users who purchase new energy vehicles in the community; the percentage of male users who purchase new energy vehicles in this market segment;
[0089] Age: The age distribution of users who purchase new energy in the community; The age distribution of users who purchase this market segment in the community; The age distribution is under 25 years old, 25-30 years old, 30-35 years old, 35-40 years old, 40-45 years old, 45-50 years old, and over 50 years old;
[0090] Whether there are children: the proportion of users who purchase new energy in the community who have children; the proportion of users who purchase new energy in this market segment in the community who have children;
[0091] Average driving time per trip: the average driving time per trip for users who purchase new energy vehicles in the community; the average driving time per trip for users who purchase new energy vehicles in this market segment;
[0092] Average driving mileage per trip: the average driving mileage per trip of users who purchase new energy in the community; the average driving mileage per trip of users who purchase new energy in this market segment;
[0093] Average daily mileage: the average daily mileage of users who purchase new energy in the community; the average daily mileage of users who purchase new energy in this market segment;
[0094] Charging start SOC: the average charging start SOC of users who purchase new energy in the community; the average charging start SOC of users who purchase new energy in this market segment in the community;
[0095] Endurance commuting times: endurance / commuting distance, summarizing the number of users who purchase new energy in the community and the number of users who purchase new energy in the market segment;
[0096] The proportion of fast charging times: fast charging times / total charging times in the community, which summarizes the users who purchase new energy in the community and the users who purchase new energy in the community in this market segment;
[0097] Number of full-charge endurance tests: sum of mileage traveled / sum of soc consumed*100% / average mileage, which is summarized separately for users who purchase new energy vehicles in the community and users who purchase new energy vehicles in this market segment in the community.
[0098] Data exploration is conducted on community purchasing potential information, community infrastructure capacity information, and community community information. Exploratory data descriptive statistical analysis and correlation analysis are performed based on community data, vehicle data, charging station data, population data, etc., to analyze the distribution differences in community purchasing potential and infrastructure capacity configuration, and whether there is a correlation between community community matching and the number of existing new energy vehicles in the community.
[0099] In a preferred embodiment, after the first type of indicators are standardized, the community purchase potential index and the infrastructure capacity configuration index are calculated based on the coefficient of variation of the single indicator, including:
[0100] The first category of indicators is subdivided into the third category of indicators and the fourth category of indicators according to the preset indicator classification standards;
[0101] Based on the third type of indicators, the purchase potential index of the community is calculated by the first type of calculation method. Based on the fourth type of indicators, the infrastructure capacity configuration index is calculated by the first type of calculation method.
[0102] Among them, the community purchase potential index calculated based on the third type of indicators through the first type of calculation method includes:
[0103] The range normalization method is used to find the maximum and minimum values of a single indicator in the third category of indicators, and the indicators are normalized to the maximum and minimum values:
[0104]
[0105] in, is the corresponding j-th indicator after standardization. is the jth index before standardization. represents the largest j-th indicator determined among all the communities in the specified city, It represents the smallest j-th indicator determined among all the communities in the specified city;
[0106] Calculate the coefficient of variation for a single indicator:
[0107]
[0108] in, represents the coefficient of variation of the j-th indicator, represents the standard deviation of the jth indicator, represents the average value of the jth indicator;
[0109] Calculate the community purchasing potential index:
[0110]
[0111] in, Indicates the purchasing potential index of the community. represents the correction coefficient of the jth indicator;
[0112] The infrastructure capacity configuration index calculated based on the fourth type of indicators using the first type of calculation method includes:
[0113] The range normalization method is used to find the maximum and minimum values of a single indicator in the fourth category of indicators, and the indicators are normalized to the maximum and minimum values:
[0114]
[0115] in, is the corresponding i-th indicator after standardization, is the corresponding i-th indicator before standardization, represents the largest i-th indicator determined among all the communities in a specified city, It represents the smallest i-th indicator determined among all the communities in the specified city;
[0116] Calculate the coefficient of variation for a single indicator:
[0117]
[0118] in, represents the coefficient of variation of the i-th indicator, represents the standard deviation of the i-th indicator, represents the average value of the i-th indicator;
[0119] Calculate the infrastructure capacity configuration index:
[0120]
[0121] in, represents the infrastructure capacity configuration index, Represents the correction coefficient of the i-th indicator.
[0122] The working principle and beneficial effects of the above technical solution are as follows: the residential advertising location index is based on three parts: the purchase potential of the community, the configuration of infrastructure capacity, and the matching degree of the community community. The purchase potential of the community and the configuration of infrastructure capacity are calculated based on the coefficient of variation of indicators in different cities. The smaller the coefficient of variation, the smaller the data fluctuation and the relatively stable data distribution. The smaller the impact of this indicator on the final index, the smaller the impact of this indicator on the final index, thereby realizing the differentiated calculation of the two indexes.
[0123] In a preferred embodiment, determining a plurality of second-category indicators based on the first-category information, inputting a pre-trained learning model into the second-category indicators and outputting a community matching index includes:
[0124] The second category of indicators is subdivided into the fifth category of indicators and the sixth category of indicators, and the fifth category of indicators is used as the independent variable and the sixth category of indicators is used as the dependent variable;
[0125] The feature factors are screened by random forest, and the importance of features is ranked according to the contribution of different features to the data set, and the indicator factors with higher importance are selected;
[0126] According to different cities, 70% of the data are selected as the experimental set and 30% as the test set;
[0127] Different regression algorithms are used to train the target value, and the model results are evaluated by mean square error, root mean square error, and mean absolute error. The underfitting and overfitting models are optimized after adjusting the parameters.
[0128] Select the optimal model, put the corresponding market segmentation data indicators into the model, and output the community matching index.
[0129] The working principle and beneficial effects of the above technical solution are as follows: Since the community matching degree of a single market segment is directly calculated and the data is sparse, machine learning modeling is first performed on the new energy vehicle data of the entire community to evaluate the score of a single market segment in the community based on the market segment.
[0130] In a preferred embodiment, according to the community purchase potential index, infrastructure capacity configuration index and community community matching index, the advertising recommendation index of the corresponding community is calculated based on the preset weight coefficient, including:
[0131] Obtain the corresponding weight coefficients of the preset community purchase potential index, infrastructure capacity configuration index and community community matching index;
[0132] The advertising recommendation index of the community is calculated by the following formula:
[0133]
[0134] in, Indicates the purchase potential index of the community. represents the infrastructure capacity configuration index, represents the community matching index, , , They represent the weight coefficients of the community purchasing potential index, infrastructure capacity configuration index and community community matching index respectively. It represents the minimum value of the purchase potential index of all communities in each city. It represents the maximum value of the purchase potential index of all communities in each city. It represents the minimum value of the infrastructure capacity configuration index of all communities in each city. It represents the maximum value of the infrastructure capacity configuration index of all communities in each city. It represents the minimum value of the community matching index of all communities in each city. It represents the maximum value of the community matching index of all communities in each city, so as to finally realize the weighted calculation of the community's advertising recommendation index.
[0135] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made in form and details without departing from the scope defined by the claims of the present invention.
Claims
1. A method for recommending residential automobile advertising sites based on Internet of Vehicles data, characterized in that: include: Get the first category information of multiple residential communities in a specified city; Based on the first type of information, multiple first type indicators are determined, and after the first type indicators are standardized, the coefficient of variation of the single indicator is used to calculate the community purchase potential index and the infrastructure capacity configuration index; Determine a plurality of second-category indicators based on the first-category information, input a pre-trained learning model into the second-category indicators and output a community matching index; According to the community purchasing potential index, infrastructure capacity configuration index and community community matching index, the advertising recommendation index of the corresponding community is calculated based on the preset weight coefficient; Determine the advertising recommendation index corresponding to each of the residential communities in the designated city, and select the recommended addresses in order of the advertising recommendation index; The first type of information includes community purchase potential information, community infrastructure capacity information and community community information, among which: The purchase potential information of the community includes the number of households in the community, the average price of vehicles in the community, the total number of new energy vehicles, the number of new energy vehicles in the same market segment, the average age of new energy vehicles purchased, the plot ratio of the community, the greening rate of the community, the distance between the community and the subway station, the number of households in the community, the number of buildings in the community, and the number of parking spaces in the community; The community infrastructure capacity information includes the proportion of private charging piles, the distance to the nearest public charging pile, the proportion of fast charging piles in the area, the car-to-pile ratio, the convenience of charging, and the waiting time for charging; Community information includes gender, age, whether there are children, average driving time per trip, average driving mileage per trip, average daily driving mileage, charging start SOC, endurance / commuting distance, proportion of fast charging times, and full-charge endurance times; The first type of indicators are standardized and the community purchase potential index and infrastructure capacity configuration index are calculated based on the coefficient of variation of the single indicator, respectively, including: The first category of indicators is subdivided into the third category of indicators and the fourth category of indicators according to the preset indicator classification standards; Based on the third type of indicators, the purchase potential index of the community is calculated by the first type of calculation method. Based on the fourth type of indicators, the infrastructure capacity configuration index is calculated by the first type of calculation method. Among them, the community purchase potential index calculated based on the third type of indicators through the first type of calculation method includes: The range normalization method is used to find the maximum and minimum values of a single indicator in the third category of indicators, and the indicators are normalized to the maximum and minimum values: Among them, X j ′ is the corresponding j-th index after standardization, X j is the jth index before standardization, X j,max represents the largest j-th index among all the communities in a given city, X j,min It represents the smallest j-th indicator determined among all the communities in the specified city; Calculate the coefficient of variation for a single indicator: Among them, C j represents the coefficient of variation of the jth indicator, σ j represents the standard deviation of the jth indicator, μ j represents the average value of the jth indicator; Calculate the community purchasing potential index: Among them, index_buy represents the purchase potential index of the community, K j represents the correction coefficient of the jth indicator; The infrastructure capacity configuration index calculated based on the fourth type of indicators using the first type of calculation method includes: The range normalization method is used to find the maximum and minimum values of a single indicator in the fourth category of indicators, and the indicators are normalized to the maximum and minimum values: Among them, X i ′ is the corresponding index after standardization, X i is the corresponding index before standardization, X i,max represents the largest i-th index among all the communities in a given city, X i,min It represents the smallest i-th indicator determined among all the communities in the specified city; Calculate the coefficient of variation for a single indicator: Among them, C i represents the coefficient of variation of the i-th indicator, σ i represents the standard deviation of the i-th indicator, μ i represents the average value of the i-th indicator; Calculate the infrastructure capacity configuration index: Among them, index_bcapacity represents the infrastructure capacity configuration index, K i Represents the correction coefficient of the i-th indicator.
2. The method for recommending residential automobile advertising sites based on Internet of Vehicles data according to claim 1, characterized in that: The determining of a plurality of second-category indicators based on the first-category information, inputting a pre-trained learning model into the second-category indicators and outputting a community matching index comprises: The second category of indicators is subdivided into the fifth category of indicators and the sixth category of indicators, and the fifth category of indicators is used as the independent variable and the sixth category of indicators is used as the dependent variable; The feature factors are screened by random forest, and the importance of features is ranked according to the contribution of different features to the data set, and the indicator factors with higher importance are selected; According to different cities, 70% of the data are selected as the experimental set and 30% as the test set; Different regression algorithms are used to train the target value, and the model results are evaluated by the mean square error, root mean square error, and mean absolute error. The underfitting model is optimized after adjusting the parameters. Select the optimal model, put the corresponding market segmentation data indicators into the model, and output the community matching index.
3. The method for recommending residential automobile advertising sites based on Internet of Vehicles data according to claim 1, characterized in that: The advertisement recommendation index of the corresponding community is calculated based on the preset weight coefficient according to the community purchase potential index, the infrastructure capacity configuration index and the community community matching index, including: Obtain the corresponding weight coefficients of the preset community purchase potential index, infrastructure capacity configuration index and community community matching index; The advertising recommendation index of the community is calculated by the following formula: Among them, index1 represents the community purchase potential index, index2 represents the infrastructure capacity configuration index, index3 represents the community community matching index, Z1, Z2, and Z3 represent the weight coefficients corresponding to the community purchase potential index, the infrastructure capacity configuration index, and the community community matching index, respectively. Index1min represents the minimum value of the community purchase potential index of all communities in each city, and index1max represents the maximum value of the community purchase potential index of all communities in each city. Index2min represents the minimum value of the infrastructure capacity configuration index of all communities in each city, and index2max represents the maximum value of the infrastructure capacity configuration index of all communities in each city. Index3min represents the minimum value of the community community matching index of all communities in each city, and index3max represents the maximum value of the community community matching index of all communities in each city.
