Intelligent Recommendation Method and System for Vehicle Owner Information Services Based on Big Data

Abstract

The present invention relates to the technical field of big data intelligent recommendation, and particularly to an intelligent recommendation method and system for vehicle owner information services based on big data, including: establishing a vehicle owner information service matching mechanism according to the pre-acquired vehicle owner service demand data, vehicle owner basic information data, and vehicle owner service registration data, and obtaining the vehicle owner information service feature vector; calculating the vehicle owner service information similarity through the cosine similarity - Pearson correlation coefficient algorithm based on the basic service data and the vehicle owner information service feature vector; based on the vehicle owner service information similarity, solving from bottom to top through a constructed multi-layer model to obtain the optimal value of vehicle owner information service recommendation; the present invention analyzes the needs of vehicle owners and formulates corresponding vehicle owner information service recommendations according to the vehicle owner needs, which can not only improve service efficiency and information recommendation accuracy, but also improve customer satisfaction, thereby increasing revenue.

Description

Technical Field

[0001] The present invention relates to the technical field of big data intelligent recommendation, and particularly relates to an intelligent recommendation method and system for vehicle owner information services based on big data. Background Art

[0002] Currently, the four main service models in the automotive sales market, namely brand 4S stores, authorized service stations, large repair shops, and private individual businesses, can no longer fully meet the needs of vehicle owners. It is necessary to analyze and summarize the types of vehicle owners' needs in detail, and there is an urgent need to change from passive service to active service.

[0003] Now, the value and potential of automotive services are far higher than those of the automotive manufacturing chain. Moreover, the vehicle ownership is showing a continuous growth trend within a certain period. The profit margin of automotive production and sales will gradually decrease, and the importance of after-sales automotive services becomes even more indispensable. Since the automotive service industry is currently in its infancy and the entire automotive service industry is relatively chaotic, comprehensive services and value-added services are seriously lacking. Traditional intelligent recommendation methods for vehicle owner information services also have problems such as inaccurate recommendations, low recommendation efficiency, and reduced automotive sales volume and unsatisfactory economic benefits. Summary of the Invention

[0004] To overcome the deficiencies of the above-mentioned prior art, the present invention provides an intelligent recommendation method for vehicle owner information services based on big data, including:

[0005] Establishing a vehicle owner information service matching mechanism according to the pre-acquired vehicle owner service demand data, vehicle owner basic information data, and vehicle owner service registration data, and obtaining the vehicle owner information service feature vector;

[0006] Calculating the vehicle owner service information similarity according to the basic service data and the vehicle owner information service feature vector through the cosine similarity - Pearson correlation coefficient algorithm;

[0007] Based on the vehicle owner service information similarity, solving from bottom to top by using the constructed multi-layer model to obtain the optimal value of vehicle owner information service recommendation;

[0008] Performing intelligent recommendation of vehicle owner information services according to the optimal value of vehicle owner information service recommendation through the fuzzy prediction method;

[0009] Among them, the vehicle owner information service matching mechanism includes: a user role matching mechanism, a user behavior habit matching mechanism, and a resource service matching mechanism;

[0010] The multi-layer model includes: a lower layer model constructed based on network optimization, a middle layer model based on intelligent improvement, and an upper layer model based on fast recommendation.

[0011] Preferably, a vehicle owner information service matching mechanism is established based on the pre-acquired vehicle owner service demand data, vehicle owner information data, and vehicle owner service registration data, and a vehicle owner information service feature vector is obtained, including:

[0012] Perform data analysis on the pre-acquired vehicle owner service demand data through big data technology to obtain vehicle owner service preference data and service cycle data;

[0013] Map and match the vehicle owner service preference data and service cycle data according to the vehicle owner information data and vehicle owner service registration data to determine the information service matching relationship;

[0014] Establish a vehicle owner information service matching mechanism through the information matching relationship, and determine matching parameters through the vehicle owner information service matching mechanism;

[0015] Perform data cleaning and data conversion operations on the matching parameters in sequence to form a matching matrix;

[0016] Obtain the vehicle owner information service feature vector according to the matching matrix;

[0017] Among them, the vehicle owner service demand data includes: maintenance service demand data, repair service demand data, value-added service demand data, vehicle agency demand data, vehicle performance demand data, and other value-added service demands for vehicle after-sales;

[0018] The vehicle owner basic information data includes: vehicle model, vehicle owner gender, vehicle owner age, vehicle owner occupation, and license plate number;

[0019] The vehicle owner service registration data includes: user account, user car purchase time, user car purchase number, user phone, and dealership information.

[0020] Preferably, the vehicle owner service information similarity is calculated through the cosine similarity-Pearson correlation coefficient algorithm based on the basic service data and the vehicle owner information service feature vector, including:

[0021] Establish a vehicle owner profile feature set based on the basic service data and the vehicle owner information service feature vector, and score the vehicle owner information service feature vector in the vehicle owner profile feature set to obtain a scoring result;

[0022] Based on the scoring result, generate a scoring matrix according to the scoring type and calculate the index value of the scoring matrix;

[0023] According to the index value, perform nearest neighbor search through the cosine similarity-Pearson correlation coefficient algorithm to obtain the vehicle owner service information similarity;

[0024] Among them, the basic service data is obtained through a big data platform;

[0025] The basic service data includes: parking service data, online car-hailing service data, route service data, and road condition service data.

[0026] Preferably, based on the similarity of the vehicle owner service information, a bottom-up solution is adopted through a pre-constructed multi-layer model to obtain the optimal value of the vehicle owner information service recommendation, including:

[0027] Based on the similarity of the vehicle owner service information, the lower-layer model is solved by the enumeration method to obtain the optimized value of the vehicle owner information service;

[0028] Based on the optimized value of the vehicle owner information, the middle-layer model is solved by the elitist retention genetic algorithm to obtain the intelligent optimized value of the vehicle owner information service;

[0029] Based on the intelligent optimized value of the vehicle owner service information, the upper-layer model is solved by the cone programming method to obtain the optimal value of the vehicle owner information service recommendation.

[0030] Preferably, the construction process of the multi-layer model is as follows:

[0031] According to the basic service data, vehicle owner service demand data, and network optimization parameters, the objective function of the lower-layer model is constructed;

[0032] Set the constraint conditions of the lower-layer model to obtain the lower-layer model;

[0033] According to the basic service data, vehicle owner service demand data, and intelligent service parameters, the objective function of the middle-layer model is constructed;

[0034] Set the constraint conditions of the middle-layer model to obtain the middle-layer model;

[0035] According to the basic service data, vehicle owner service demand data, and fast recommendation parameters, the objective function of the upper-layer model is constructed;

[0036] Set the constraint conditions of the upper-layer model to obtain the upper-layer model;

[0037] According to the lower-layer model, middle-layer model, and upper-layer model, model integration is performed to obtain the multi-layer model.

[0038] Preferably, the constraint conditions of the lower-layer model include: maximum capacity constraint for network information data transmission, constraint on the amount of rating result feedback, and maximum value constraint on the similarity of vehicle owner service information;

[0039] The constraint conditions of the middle-layer model include: maximum threshold constraint for service intelligence, constraint on the retrieval speed of basic service data, and constraint on the range of vehicle transaction quantities;

[0040] The upper-layer model constraint conditions include: information service recommendation time constraint, maximum intelligent recommendation accuracy constraint, and owner information service intelligent recommendation satisfaction constraint.

[0041] Preferably, the objective function calculation formula of the upper-layer model is as follows: In the formula, represents the optimal value of the owner information service recommendation; represents the intelligent optimization value of the owner information service; represents the basic service data; represents the quick recommendation parameter; represents the similarity parameter; represents the scoring result; represents the index value of the scoring matrix; represents the total number of basic service data; represents the number of basic service data; represents the number of scoring types; represents the total number of scoring types.

[0042] Preferably, the objective function calculation formula of the middle-layer model is as follows: In the formula, represents the intelligent optimization value of the owner information service; represents the optimization value of the owner information service; represents the total number of basic service data; represents the number of basic service data; represents the basic service data; represents the weight value of the scoring matrix; represents the intelligent service parameter.

[0043] Preferably, the objective function calculation formula of the lower-layer model is as follows: In the formula, represents the optimization value of the owner information service; represents the vehicle information service feature vector; represents the weight value of the scoring matrix; represents the number of vehicle owner information service feature vectors; represents the total number of vehicle owner information service feature vectors; L represents the network optimization parameter; represents the index value of the scoring matrix; represents the similarity parameter; represents the number of scoring types; represents the total number of scoring types; represents the scoring result.

[0044] Based on the same inventive concept, the present invention also provides a big data-based intelligent recommendation system for vehicle owner information services, including:

[0045] A feature vector acquisition module, configured to establish a vehicle owner information service matching mechanism according to pre-acquired vehicle owner service demand data, vehicle owner basic information data, and vehicle owner service registration data, and acquire a vehicle owner information service feature vector;

[0046] A service information calculation module, configured to calculate a vehicle owner service information similarity according to basic service data and the vehicle owner information service feature vector through a cosine similarity-Pearson correlation coefficient algorithm;

[0047] A recommended optimal value acquisition module, configured to, based on the vehicle owner service information similarity, solve bottom-up through a pre-constructed multi-layer model to acquire a recommended optimal value of the vehicle owner information service;

[0048] An intelligent recommendation module, configured to perform intelligent recommendation of vehicle owner information services according to the recommended optimal value of the vehicle owner information service through a fuzzy prediction method;

[0049] Wherein, the vehicle owner information service matching mechanism includes: a user role matching mechanism, a user behavior habit matching mechanism, and a resource service matching mechanism;

[0050] The multi-layer model includes: a lower layer model constructed based on network optimization, a middle layer model based on intelligent improvement, and an upper layer model based on fast recommendation.

[0051] Preferably, the feature vector acquisition module is specifically configured to:

[0052] Perform data analysis on pre-acquired vehicle owner service demand data through big data technology to acquire vehicle owner service preference data and service cycle data;

[0053] Map and match the vehicle owner service preference data and service cycle data according to vehicle owner information data and vehicle owner service registration data to determine an information service matching relationship;

[0054] Establish a vehicle owner information service matching mechanism through the information matching relationship, and determine matching parameters through the vehicle owner information service matching mechanism;

[0055] Perform data cleaning and data conversion operations on the matching parameters in sequence to form a matching matrix;

[0056] Acquire a vehicle owner information service feature vector according to the matching matrix.

[0057] Wherein, the vehicle owner service demand data includes: maintenance service demand data, repair service demand data, value-added service demand data, vehicle agency demand data, vehicle performance demand data, and other value-added service demand for vehicle after-sales service;

[0058] The basic information data of the vehicle owner includes: vehicle model, gender of the vehicle owner, age of the vehicle owner, occupation of the vehicle owner, and license plate number;

[0059] The service registration data of the vehicle owner includes: user account, vehicle purchase time of the user, vehicle purchase number of the user, user phone number, and dealership information.

[0060] Preferably, the service information calculation module is specifically used for:

[0061] Based on the basic service data and the service feature vector of the vehicle owner information, establish a feature set of the vehicle owner file, and score the service feature vector of the vehicle owner information in the feature set of the vehicle owner file to obtain a scoring result;

[0062] Based on the scoring result, generate a scoring matrix according to the scoring type, and calculate the index value of the scoring matrix;

[0063] According to the index value, perform a nearest neighbor search through the cosine similarity - Pearson correlation coefficient algorithm to obtain the similarity of the vehicle owner service information;

[0064] Among them, the basic service data is obtained through a big data platform;

[0065] The basic service data includes: parking service data, online car-hailing service data, route service data, and road condition service data.

[0066] Preferably, the recommended optimal value acquisition module is specifically used for:

[0067] Based on the similarity of the vehicle owner service information, solve the lower-layer model through the enumeration method to obtain the optimized value of the vehicle owner information service;

[0068] Based on the optimized value of the vehicle owner information, solve the middle-layer model through the elitist retention genetic algorithm to obtain the intelligent optimized value of the vehicle owner information service;

[0069] Based on the intelligent optimized value of the vehicle owner service information, solve the upper-layer model through the cone programming method to obtain the recommended optimal value of the vehicle owner information service.

[0070] Preferably, the construction process of the multi-layer model in the recommended optimal value acquisition module is as follows:

[0071] According to the basic service data, vehicle owner service demand data, and network optimization parameters, construct the objective function of the lower-layer model;

[0072] Set the constraint conditions of the lower-layer model to obtain the lower-layer model;

[0073] According to the basic service data, vehicle owner service demand data, and intelligent service parameters, construct the objective function of the middle-layer model;

[0074] Set the constraints for the middle-layer model to obtain the middle-layer model;

[0075] Construct the objective function of the upper-layer model according to the basic service data, the owner service demand data, and the quick recommendation parameters;

[0076] Set the constraints for the upper-layer model to obtain the upper-layer model;

[0077] Perform model integration according to the lower-layer model, the middle-layer model, and the upper-layer model to obtain a multi-layer model.

[0078] Preferably, the constraints for the lower-layer model in the recommended optimal value acquisition module include: the maximum capacity constraint for network information data transmission, the constraint on the amount of feedback of the scoring result, and the maximum similarity constraint for the owner service information;

[0079] The constraints for the middle-layer model include: the maximum threshold constraint for service intelligence, the constraint on the retrieval speed of basic service data, and the constraint on the range of vehicle transaction quantities;

[0080] The constraints for the upper-layer model include: the constraint on the information service recommendation time, the maximum accuracy constraint for intelligent recommendation, and the satisfaction constraint for the intelligent recommendation of the owner information service.

[0081] Preferably, the calculation formula for the objective function of the upper-layer model in the recommended optimal value acquisition module is as follows: In the formula, represents the recommended optimal value of the owner information service; represents the intelligent optimization value of the owner information service; represents the basic service data; represents the quick recommendation parameter; represents the similarity parameter; represents the scoring result; represents the index value of the scoring matrix; represents the total number of basic service data; represents the number of basic service data; represents the number of scoring types; represents the total number of scoring types.

[0082] Preferably, the calculation formula for the objective function of the middle-layer model in the recommended optimal value acquisition module is as follows: In the formula, represents the intelligent optimization value of the owner information service; represents the optimization value of the owner information service; represents the total number of basic service data; represents the number of basic service data; represents the basic service data; represents the weight value of the scoring matrix; Represents intelligent service parameters.

[0083] Preferably, the objective function calculation formula of the lower layer model in the recommended optimal value acquisition module is as follows: In the formula, Represents the optimization value of the vehicle owner information service; Represents the feature vector of the vehicle number information service; Represents the weight value of the rating matrix; Represents the number of feature vectors of the vehicle owner information service; Represents the total number of feature vectors of the vehicle owner information service; L represents the network optimization parameter; Represents the index value of the rating matrix; Represents the similarity parameter; Represents the number of rating types; Represents the total number of rating types; Represents the rating result.

[0084] Compared with the closest prior art, the beneficial effects of the present invention are as follows:

[0085] 1. The present invention provides an intelligent recommendation method for vehicle owner information services based on big data, including: establishing a vehicle owner information service matching mechanism according to the pre-acquired vehicle owner service demand data, vehicle owner basic information data, and vehicle owner service registration data to obtain the feature vector of the vehicle owner information service; calculating the similarity of the vehicle owner service information through the cosine similarity-Pearson correlation coefficient algorithm based on the basic service data and the feature vector of the vehicle owner information service; based on the similarity of the vehicle owner service information, solving from bottom to top through the constructed multi-layer model to obtain the recommended optimal value of the vehicle owner information service; and performing intelligent recommendation of the vehicle owner information service through the fuzzy prediction method; wherein, the vehicle owner information service matching mechanism includes: a user role matching mechanism, a user behavior habit matching mechanism, and a resource service matching mechanism; the multi-layer model includes: a lower layer model constructed based on network optimization, a middle layer model based on intelligent improvement, and an upper layer model based on fast recommendation. The present invention analyzes the needs of vehicle owners and formulates corresponding vehicle owner information service recommendations according to the needs of vehicle owners, which can not only improve service efficiency and information recommendation accuracy, but also improve customer satisfaction, thereby increasing revenue;

[0086] 2. The present invention calculates the similarity of the vehicle owner service information through the cosine similarity-Pearson correlation coefficient algorithm and solves from bottom to top through the constructed multi-layer model to obtain the recommended optimal value of the vehicle owner information service, with a smaller recommended error and more intelligent, and has an ideal recommendation effect. Description of the Drawings

[0087] Figure 1Flowchart of the intelligent recommendation method for vehicle owner information service based on big data provided by the present invention;

[0088] Figure 2 System module connection diagram of the intelligent recommendation method for vehicle owner information service based on big data provided by the present invention. Specific implementation manners

[0089] The following further elaborates on the specific implementation manners of the present invention with reference to the accompanying drawings.

[0090] In Embodiment 1, the flowchart of the intelligent recommendation method for vehicle owner information service based on big data provided by the present invention is as Figure 1 shown and includes:

[0091] Step 1: Establish a vehicle owner information service matching mechanism according to the pre-obtained vehicle owner service demand data, vehicle owner basic information data, and vehicle owner service registration data, and obtain the vehicle owner information service feature vector;

[0092] Step 2: Calculate the vehicle owner service information similarity through the cosine similarity - Pearson correlation coefficient algorithm based on the basic service data and the vehicle owner information service feature vector;

[0093] Step 3: Based on the vehicle owner service information similarity, solve from bottom to top through the constructed multi-layer model to obtain the optimal value of vehicle owner information service recommendation;

[0094] Step 4: Perform intelligent recommendation of vehicle owner information service through the fuzzy prediction method according to the optimal value of vehicle owner information service recommendation;

[0095] Among them, the vehicle owner information service matching mechanism includes: user role matching mechanism, user behavior habit matching mechanism, and resource service matching mechanism;

[0096] The multi-layer model includes: a lower layer model constructed based on network optimization, a middle layer model based on intelligent improvement, and an upper layer model based on fast recommendation.

[0097] Specifically, Step 1 includes:

[0098] Perform data analysis on the pre-obtained vehicle owner service demand data through big data technology to obtain vehicle owner service preference data and service cycle data;

[0099] Map and match the vehicle owner service preference data and service cycle data according to the vehicle owner information data and vehicle owner service registration data to determine the information service matching relationship;

[0100] Establish a vehicle owner information service matching mechanism through the information matching relationship, and determine matching parameters through the vehicle owner information service matching mechanism;

[0101] Perform data cleaning and data transformation operations on the matching parameters in sequence to form a matching matrix;

[0102] Obtain the vehicle owner information service feature vector according to the matching matrix;

[0103] Among them, the vehicle owner service demand data includes: maintenance service demand data, repair service demand data, value-added service demand data, vehicle agency demand data, vehicle performance demand data, and other value-added service demands such as vehicle after-sales demand data;

[0104] The vehicle owner basic information data includes: vehicle model, vehicle owner gender, vehicle owner age, vehicle owner occupation, and license plate number;

[0105] The vehicle owner service registration data includes: user account, user car purchase time, user car purchase number, user phone number, and car dealership information.

[0106] Currently, in addition to undertaking the mainstream sales of passenger cars, automobile 4S stores also undertake the channel tasks of automobile parts supply, information feedback, and after-sales service;

[0107] 4S stores are still the main places for vehicle owners to perform vehicle repairs, maintenance, etc. Vehicle owners have not been very satisfied with its charging standards and service quality;

[0108] In addition, now automobile chain quick repair shops and roadside private individual stores have sprung up in the streets and alleys, and have formed a price advantage with charging standards lower than those of 4S stores. Therefore, when the vehicle exceeds the warranty period, the loss of 4S store customers is particularly serious;

[0109] The large loss of customers directly leads to the loss of customer demand management objects and indirectly causes the lack of demand management. Therefore, analyzing the needs of vehicle owners and formulating corresponding vehicle owner information service recommendations according to the needs of vehicle owners can not only improve service efficiency but also improve customer satisfaction, thereby increasing revenue.

[0110] The vehicle owner service preference data and service cycle data obtained by the present invention through analyzing the vehicle owner service demand data are stored in the database, and mapping and matching are performed according to the vehicle owner service registration data and vehicle owner information data. Among them, the vehicle owner registration data and vehicle owner information data have a one-to-one mapping relationship;

[0111] The vehicle owner registration data and vehicle owner information data respectively have a one-to-many mapping relationship with the vehicle owner service preference data;

[0112] The vehicle owner information data and the service cycle data have a one-to-one time-limited update mapping relationship. The service cycle is obtained by recording and statistics each time the vehicle owner makes a service request and is updated with the number of vehicle owner services;

[0113] According to the above mapping correspondence, the vehicle owner information service matching mechanism established in the database can perform preference screening by inputting the keywords of each mapping data in the information service matching relationship;

[0114] Performing data cleaning and data conversion operations on the matching parameters in sequence to form a matching matrix can eliminate the interference of irrelevant data and thus ensure the accuracy of the data.

[0115] Specifically, step 2 includes:

[0116] According to the basic service data and the vehicle owner information service feature vector, establish a vehicle owner profile feature set, and score the vehicle owner information service feature vectors in the vehicle owner profile feature set to obtain a scoring result;

[0117] Based on the scoring result, generate a scoring matrix according to the scoring type and calculate the index value of the scoring matrix;

[0118] Among them, the calculation formula of the scoring matrix is as follows: In the formula, represents the scoring result; R represents the scoring score corresponding to each user vehicle owner U for different scoring types; U represents the user vehicle owner; P represents the vehicle owner information service feature vector; k represents the number of vehicle owner information service feature vectors; d represents the number of users;

[0119] Among them, the larger the R value, the higher the satisfaction of the user with the service item;

[0120] According to the index value, perform a nearest neighbor search through the cosine similarity-Pearson correlation coefficient algorithm to obtain the vehicle owner service information similarity;

[0121] Among them, the basic service data is obtained through a big data platform;

[0122] The basic service data includes: parking service data, online car-hailing service data, route service data, and road condition service data.

[0123] Among them, common similarity calculation methods include Euclidean distance method, Manhattan distance method, Minkowski distance method, cosine similarity method, Pearson correlation coefficient method, improved cosine similarity-Pearson correlation coefficient method, and Tanimoto coefficient method, etc.;

[0124] In the present invention, calculating the vehicle owner service information similarity through the improved cosine similarity-Pearson correlation coefficient method can have an obvious effect on the processing of well-formatted preference data;

[0125] Among them, the calculation formula of the vehicle owner service information similarity is as follows: In the formula, Represents the similarity of vehicle owner service information; Represents the score values corresponding to different rating types for the vehicle owners of d users, etc.; Represents the index value of the rating matrix; Represents the score values corresponding to different rating types for the k-th vehicle owner information service feature vector.

[0126] Specifically, step 3 includes:

[0127] Based on the similarity of vehicle owner service information, solve the lower-layer model by the enumeration method to obtain the optimized value of vehicle owner information service;

[0128] Based on the optimized value of vehicle owner information, solve the middle-layer model by the elitist retention genetic algorithm to obtain the intelligent optimized value of vehicle owner information service;

[0129] Based on the intelligent optimized value of vehicle owner service information, solve the upper-layer model by the cone programming method to obtain the recommended optimal value of vehicle owner information service.

[0130] The construction process of the multi-layer model is as follows:

[0131] According to the basic service data, vehicle owner service demand data, and network optimization parameters, construct the objective function of the lower-layer model;

[0132] Set the constraint conditions of the lower-layer model to obtain the lower-layer model;

[0133] According to the basic service data, vehicle owner service demand data, and intelligent service parameters, construct the objective function of the middle-layer model;

[0134] Set the constraint conditions of the middle-layer model to obtain the middle-layer model;

[0135] According to the basic service data, vehicle owner service demand data, and fast recommendation parameters, construct the objective function of the upper-layer model;

[0136] Set the constraint conditions of the upper-layer model to obtain the upper-layer model;

[0137] Integrate the lower-layer model, middle-layer model, and upper-layer model to obtain the multi-layer model.

[0138] The constraint conditions of the lower-layer model include: the maximum capacity constraint of network information data transmission, the constraint of the feedback quantity of rating results, and the maximum value constraint of the similarity of vehicle owner service information;

[0139] The constraint conditions of the middle-layer model include: the maximum threshold constraint of service intelligence, the constraint of the retrieval speed of basic service data, and the constraint of the range of vehicle transaction quantities;

[0140] The upper-layer model constraint conditions include: information service recommendation time constraint, maximum intelligent recommendation accuracy constraint, and owner information service intelligent recommendation satisfaction constraint.

[0141] The objective function calculation formula of the upper-layer model is as follows: In the formula, represents the optimal value of the owner information service recommendation; represents the intelligent optimization value of the owner information service; represents the basic service data; represents the quick recommendation parameter; represents the similarity parameter; represents the scoring result; represents the index value of the scoring matrix; represents the total number of basic service data; represents the number of basic service data; represents the number of scoring types; represents the total number of scoring types.

[0142] The objective function calculation formula of the middle-layer model is as follows: In the formula, represents the intelligent optimization value of the owner information service; represents the optimization value of the owner information service; represents the total number of basic service data; represents the number of basic service data; represents the basic service data; represents the weight value of the scoring matrix; represents the intelligent service parameter.

[0143] The objective function calculation formula of the lower-layer model is as follows: In the formula, represents the optimization value of the owner information service; represents the vehicle information service feature vector; represents the weight value of the scoring matrix; represents the number of vehicle owner information service feature vectors; represents the total number of vehicle owner information service feature vectors; L represents the network optimization parameter; represents the index value of the scoring matrix; represents the similarity parameter; represents the number of scoring types; represents the total number of scoring types; represents the scoring result.

[0144] Specifically, step 4 includes:

[0145] Recommend the optimal value according to the vehicle owner information service, and perform intelligent recommendation of the vehicle owner information service through the fuzzy prediction method;

[0146] Among them, the fuzzy prediction of the present invention refers to when the vehicle owner has no historical service record in the database or when following up on the vehicle owners with input data, calculating and analogizing the similarity based on the existing vehicle owner service demand data, vehicle owner basic information data, vehicle owner service registration data, and basic service data, and then solving the multi-layer model, and finally obtaining the optimal value of the vehicle owner information service recommendation, so as to initially predict which type of maintenance, repair, or value-added the vehicle owner belongs to, as well as the information feedback recommended based on the basic data.

[0147] Embodiment 2, the connection diagram of the intelligent recommendation system module for vehicle owner information service based on big data provided by the present invention is as Figure 2 shown, including:

[0148] A feature vector acquisition module, configured to establish a vehicle owner information service matching mechanism according to the pre-acquired vehicle owner service demand data, vehicle owner basic information data, and vehicle owner service registration data, and acquire the vehicle owner information service feature vector;

[0149] A service information calculation module, configured to calculate the vehicle owner service information similarity according to the basic service data and the vehicle owner information service feature vector through the cosine similarity - Pearson correlation coefficient algorithm;

[0150] A recommended optimal value acquisition module, configured to solve from bottom to top through a pre-constructed multi-layer model based on the vehicle owner service information similarity, and acquire the optimal value of the vehicle owner information service recommendation;

[0151] An intelligent recommendation module, configured to perform intelligent recommendation of the vehicle owner information service through the fuzzy prediction method according to the optimal value of the vehicle owner information service recommendation;

[0152] Among them, the vehicle owner information service matching mechanism includes: a user role matching mechanism, a user behavior habit matching mechanism, and a resource service matching mechanism;

[0153] The multi-layer model includes: a lower layer model constructed based on network optimization, a middle layer model based on intelligent improvement, and an upper layer model based on fast recommendation.

[0154] Specifically, the feature vector acquisition module is specifically configured to:

[0155] Perform data analysis on the pre-acquired vehicle owner service demand data through big data technology to obtain vehicle owner service preference data and service cycle data;

[0156] Map and match the vehicle owner service preference data and service cycle data according to the vehicle owner information data and vehicle owner service registration data to determine the information service matching relationship;

[0157] Establish a vehicle owner information service matching mechanism through the information matching relationship, and determine matching parameters through the vehicle owner information service matching mechanism;

[0158] Perform data cleaning and data conversion operations on the matching parameters in sequence to form a matching matrix;

[0159] Obtain the vehicle owner information service feature vector according to the matching matrix.

[0160] Among them, the vehicle owner service demand data includes: maintenance service demand data, repair service demand data, value-added service demand data, vehicle agency demand data, vehicle performance demand data, and other value-added service demands for vehicle after-sales service;

[0161] The vehicle owner basic information data includes: vehicle model, vehicle owner gender, vehicle owner age, vehicle owner occupation, and license plate number;

[0162] The vehicle owner service registration data includes: user account, user car purchase time, user car purchase number, user phone, and car dealership information.

[0163] Specifically, the service information calculation module is specifically used for:

[0164] Establish a vehicle owner file feature set according to the basic service data and the vehicle owner information service feature vector, and score the vehicle owner information service feature vector in the vehicle owner file feature set to obtain a scoring result;

[0165] Based on the scoring result, generate a scoring matrix according to the scoring type and calculate the index value of the scoring matrix;

[0166] According to the index value, perform nearest neighbor search through the cosine similarity-Pearson correlation coefficient algorithm to obtain the vehicle owner service information similarity;

[0167] Among them, the basic service data is obtained through a big data platform;

[0168] The basic service data includes: parking service data, online car-hailing service data, route service data, and road condition service data.

[0169] Specifically, the recommended optimal value acquisition module is specifically used for:

[0170] Based on the vehicle owner service information similarity, solve the lower-layer model through the enumeration method to obtain the vehicle owner information service optimization value;

[0171] Based on the optimized value of the vehicle owner information, the middle - layer model is solved by the elitist - retention genetic algorithm to obtain the intelligent optimized value of the vehicle owner information service;

[0172] Based on the intelligent optimized value of the vehicle owner service information, the upper - layer model is solved by the cone programming method to obtain the optimal recommended value of the vehicle owner information service.

[0173] The construction process of the multi - layer model in the optimal recommended value acquisition module is as follows:

[0174] According to the basic service data, vehicle owner service demand data, and network optimization parameters, the objective function of the lower - layer model is constructed;

[0175] Set the constraint conditions of the lower - layer model to obtain the lower - layer model;

[0176] According to the basic service data, vehicle owner service demand data, and intelligent service parameters, the objective function of the middle - layer model is constructed;

[0177] Set the constraint conditions of the middle - layer model to obtain the middle - layer model;

[0178] According to the basic service data, vehicle owner service demand data, and fast - recommendation parameters, the objective function of the upper - layer model is constructed;

[0179] Set the constraint conditions of the upper - layer model to obtain the upper - layer model;

[0180] According to the lower - layer model, middle - layer model, and upper - layer model, model integration is performed to obtain the multi - layer model.

[0181] The constraint conditions of the lower - layer model in the optimal recommended value acquisition module include: the maximum capacity constraint of network information data transmission, the constraint of the amount of rating result feedback, and the maximum similarity constraint of vehicle owner service information;

[0182] The constraint conditions of the middle - layer model include: the maximum threshold constraint of service intelligence, the constraint of the retrieval speed of basic service data, and the constraint of the vehicle transaction quantity range;

[0183] The constraint conditions of the upper - layer model include: the constraint of information service recommendation time, the maximum accuracy constraint of intelligent recommendation, and the satisfaction constraint of intelligent recommendation of vehicle owner information service.

[0184] The calculation formula of the objective function of the upper - layer model in the optimal recommended value acquisition module is as follows: In the formula, represents the optimal recommended value of the vehicle owner information service; represents the intelligent optimized value of the vehicle owner information service; represents the basic service data; represents the fast - recommendation parameter; represents the similarity parameter; represents the rating result; Indicates the index value of the scoring matrix; Indicates the total number of basic service data; Indicates the number of basic service data; Indicates the number of scoring types; Indicates the total number of scoring types.

[0185] The objective function calculation formula of the middle-layer model in the recommended optimal value acquisition module is as follows: In the formula, Indicates the intelligent optimization value of the vehicle owner information service; Indicates the optimization value of the vehicle owner information service; Indicates the total number of basic service data; Indicates the number of basic service data; Indicates the basic service data; Indicates the weight value of the scoring matrix; Indicates the intelligent service parameter.

[0186] The objective function calculation formula of the lower-layer model in the recommended optimal value acquisition module is as follows: In the formula, Indicates the optimization value of the vehicle owner information service; Indicates the feature vector of the vehicle information service; Indicates the weight value of the scoring matrix; Indicates the number of feature vectors of the vehicle owner information service; Indicates the total number of feature vectors of the vehicle owner information service; L represents the network optimization parameter; Indicates the index value of the scoring matrix; Indicates the similarity parameter; Indicates the number of scoring types; Indicates the total number of scoring types; Indicates the scoring result.

[0187] Specifically, the intelligent recommendation module is specifically used for:

[0188] To recommend the optimal value according to the vehicle owner information service and perform intelligent recommendation of the vehicle owner information service through the fuzzy prediction method.

[0189] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0190] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded processors, or other programmable data processing devices to generate a machine, such that the instructions executed by the processors of the computer or other programmable data processing devices generate means for implementing the functions specified in one or more of the flows Figure 1 one or more flows and / or blocks Figure 1 or means for implementing the functions specified in one or more of the blocks.

[0191] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means for implementing the functions specified in one or more of the flows Figure 1 one or more flows and / or blocks Figure 1 or means for implementing the functions specified in one or more of the blocks.

[0192] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more of the flows Figure 1 one or more flows and / or blocks Figure 1 or means for implementing the functions specified in one or more of the blocks.

[0193] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit the scope of its protection. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that after reading the present invention, various changes, modifications, or equivalent replacements can still be made to the specific implementation manners of the application. However, these changes, modifications, or equivalent replacements are all within the scope of the protection of the claims pending for the application.

Claims

1. An intelligent recommendation method for car owner information service based on big data, characterized in that: include: Establishing a vehicle owner information service matching mechanism based on the pre-acquired vehicle owner service demand data, vehicle owner basic information data and vehicle owner service registration data, and obtaining a vehicle owner information service feature vector; Calculating the similarity of the vehicle owner service information based on the basic service data and the vehicle owner information service feature vector by using a cosine similarity-Pearson correlation coefficient algorithm; Based on the similarity of the vehicle owner service information, a multi-layer model is constructed to solve the problem from bottom to top to obtain the optimal value of the vehicle owner information service recommendation; According to the optimal value of the vehicle owner information service recommendation, intelligently recommend the vehicle owner information service through a fuzzy prediction method; The vehicle owner information service matching mechanism includes: user role matching mechanism, user behavior habit matching mechanism and resource service matching mechanism; The multi-layer model includes: a lower-layer model based on network optimization, a middle-layer model based on intelligent improvement, and an upper-layer model based on fast recommendation; The step of establishing a vehicle owner information service matching mechanism based on the pre-acquired vehicle owner service demand data, vehicle owner information data and vehicle owner service registration data, and acquiring a vehicle owner information service feature vector, includes: Based on the pre-acquired car owner service demand data, data analysis is performed using big data technology to obtain the car owner's service preference data and service cycle data; Mapping and matching the vehicle owner service preference data and service cycle data according to the vehicle owner information data and the vehicle owner service registration data to determine an information service matching relationship; Establishing a vehicle owner information service matching mechanism through the information matching relationship, and determining matching parameters through the vehicle owner information service matching mechanism; Performing data cleaning and data conversion operations on the matching parameters in sequence to form a matching matrix; According to the matching matrix, obtaining a vehicle owner information service feature vector; The car owner service demand data includes: maintenance service demand data, repair service demand data, value-added service demand data, car agency demand data, car performance demand data, car after-sales demand data and other value-added service demands; The basic information data of the vehicle owner includes: vehicle model, gender, age, occupation and license plate number; The car owner service registration data includes: user account, user car purchase time, user car purchase number, user phone number and car dealer information; The construction process of the multi-layer model is as follows: Construct the objective function of the lower-level model based on basic service data, car owner service demand data and network optimization parameters; Set the constraints of the lower model to obtain the lower model; Constructing a mid-level model objective function based on the basic service data, the vehicle owner service demand data and the intelligent service parameters; Set the constraints of the middle-level model to obtain the middle-level model; Constructing an upper-level model objective function according to the basic service data, the vehicle owner service demand data and the quick recommendation parameters; Set the constraints of the upper model to obtain the upper model; Perform model integration according to the lower-level model, the middle-level model and the upper-level model to obtain a multi-layer model; The lower model constraints include: a maximum capacity constraint for network information data transmission, a constraint on the amount of rating result feedback, and a maximum similarity constraint for vehicle owner service information; The middle-level model constraints include: service intelligence maximum threshold constraint, basic service data retrieval speed constraint and vehicle transaction quantity range constraint; The upper model constraints include: information service recommendation time constraint, intelligent recommendation accuracy maximum constraint and owner information service intelligent recommendation satisfaction constraint; The objective function calculation formula of the upper model is as follows: In the formula, Indicates the optimal value recommended by the vehicle owner information service; Indicates the intelligent optimization value of the car owner information service; Represents basic service data; Indicates quick recommended parameters; represents the similarity parameter; Indicates the scoring result; Represents the indicator value of the scoring matrix; Indicates the total number of basic service data; Indicates the number of basic service data; Indicates the number of rating types; Indicates the total number of rating types; The objective function calculation formula of the middle-level model is as follows: In the formula, Indicates the intelligent optimization value of the car owner information service; Indicates the optimization value of the vehicle owner information service; Indicates the total number of basic service data; Indicates the number of basic service data; Represents basic service data; Represents the weight value of the scoring matrix; Indicates intelligent service parameters; The objective function calculation formula of the lower model is as follows: In the formula, Indicates the optimization value of the vehicle owner information service; Represents the vehicle number information service feature vector; Represents the weight value of the scoring matrix; Indicates the number of feature vectors of the car owner information service; represents the total number of feature vectors of the car owner information service; L represents the network optimization parameter; Represents the indicator value of the scoring matrix; represents the similarity parameter; Indicates the number of rating types; Indicates the total number of rating types; Indicates the scoring result.

2. The intelligent recommendation method for car owner information service based on big data as claimed in claim 1, characterized in that: The calculating the similarity of the vehicle owner service information according to the basic service data and the vehicle owner information service feature vector by using a cosine similarity-Pearson correlation coefficient algorithm includes: Establishing a vehicle owner profile feature set according to the basic service data and the vehicle owner information service feature vector, and scoring the vehicle owner information service feature vector in the vehicle owner profile feature set to obtain a scoring result; Based on the scoring results, a scoring matrix is ​​generated according to the scoring type, and the index values ​​of the scoring matrix are calculated; According to the index value, a nearest neighbor search is performed using a cosine similarity-Pearson correlation coefficient algorithm to obtain the similarity of the vehicle owner's service information; Wherein, the basic service data is obtained through a big data platform; The basic service data includes: parking service data, online booking service data, route service data and road condition service data.

3. The intelligent recommendation method for car owner information service based on big data as claimed in claim 1, characterized in that: The method of obtaining the optimal value of the vehicle owner information service recommendation based on the vehicle owner service information similarity by using a pre-built multi-layer model and bottom-up solving, includes: Based on the similarity of the vehicle owner service information, the lower layer model is solved by enumeration method to obtain the vehicle owner information service optimization value; Based on the optimized value of the vehicle owner information, the middle-level model is solved by an elite-reserved genetic algorithm to obtain an intelligent optimized value of the vehicle owner information service; Based on the intelligent optimization value of the car owner service information, the upper model is solved by the cone programming method to obtain the optimal value of the car owner information service recommendation.

4. Intelligent recommendation system for car owner information service based on big data, characterized by: include: A feature vector acquisition module is used to establish a vehicle owner information service matching mechanism based on the pre-acquired vehicle owner service demand data, vehicle owner basic information data and vehicle owner service registration data, and acquire a vehicle owner information service feature vector; A service information calculation module, used to calculate the similarity of the vehicle owner service information according to the basic service data and the vehicle owner information service feature vector by using a cosine similarity-Pearson correlation coefficient algorithm; A recommended optimal value acquisition module, used for obtaining the recommended optimal value of the vehicle owner information service based on the vehicle owner service information similarity by solving the problem from bottom to top through a pre-built multi-layer model; An intelligent recommendation module, used to recommend the optimal value of the vehicle owner information service and to make intelligent recommendations for the vehicle owner information service through a fuzzy prediction method; The vehicle owner information service matching mechanism includes: user role matching mechanism, user behavior habit matching mechanism and resource service matching mechanism; The multi-layer model includes: a lower-layer model based on network optimization, a middle-layer model based on intelligent improvement, and an upper-layer model based on fast recommendation; The step of establishing a vehicle owner information service matching mechanism based on the pre-acquired vehicle owner service demand data, vehicle owner information data and vehicle owner service registration data, and acquiring a vehicle owner information service feature vector, includes: Based on the pre-acquired car owner service demand data, data analysis is performed using big data technology to obtain the car owner's service preference data and service cycle data; Mapping and matching the vehicle owner service preference data and service cycle data according to the vehicle owner information data and the vehicle owner service registration data to determine an information service matching relationship; Establishing a vehicle owner information service matching mechanism through the information service matching relationship, and determining matching parameters through the vehicle owner information service matching mechanism; Performing data cleaning and data conversion operations on the matching parameters in sequence to form a matching matrix; According to the matching matrix, obtaining a vehicle owner information service feature vector; The car owner service demand data includes: maintenance service demand data, repair service demand data, value-added service demand data, car agency demand data, car performance demand data, car after-sales demand data and other value-added service demands; The basic information data of the vehicle owner includes: vehicle model, gender, age, occupation and license plate number; The car owner service registration data includes: user account, user car purchase time, user car purchase number, user phone number and car dealer information; The construction process of the multi-layer model is as follows: Construct the objective function of the lower-level model based on basic service data, car owner service demand data and network optimization parameters; Set the constraints of the lower model to obtain the lower model; Constructing a mid-level model objective function based on the basic service data, the vehicle owner service demand data and the intelligent service parameters; Set the constraints of the middle-level model to obtain the middle-level model; Constructing an upper-level model objective function according to the basic service data, the vehicle owner service demand data and the quick recommendation parameters; Set the constraints of the upper model to obtain the upper model; Perform model integration according to the lower-level model, the middle-level model and the upper-level model to obtain a multi-layer model; The lower model constraints include: a maximum capacity constraint for network information data transmission, a constraint on the amount of rating result feedback, and a maximum similarity constraint for vehicle owner service information; The middle-level model constraints include: service intelligence maximum threshold constraint, basic service data retrieval speed constraint and vehicle transaction quantity range constraint; The upper model constraints include: information service recommendation time constraint, intelligent recommendation accuracy maximum constraint and owner information service intelligent recommendation satisfaction constraint; The objective function calculation formula of the upper model is as follows: In the formula, Indicates the optimal value recommended by the vehicle owner information service; Indicates the intelligent optimization value of the car owner information service; Represents basic service data; Indicates quick recommended parameters; represents the similarity parameter; Indicates the scoring result; Represents the indicator value of the scoring matrix; Indicates the total number of basic service data; Indicates the number of basic service data; Indicates the number of rating types; Indicates the total number of rating types; The objective function calculation formula of the middle-level model is as follows: In the formula, Indicates the intelligent optimization value of the car owner information service; Indicates the optimization value of the vehicle owner information service; Indicates the total number of basic service data; Indicates the number of basic service data; Represents basic service data; Represents the weight value of the scoring matrix; Indicates intelligent service parameters; The objective function calculation formula of the lower model is as follows: In the formula, Indicates the optimization value of the vehicle owner information service; Represents the vehicle number information service feature vector; Represents the weight value of the scoring matrix; Indicates the number of feature vectors of car owner information service; represents the total number of feature vectors of the car owner information service; L represents the network optimization parameter; Represents the indicator value of the scoring matrix; represents the similarity parameter; Indicates the number of rating types; Indicates the total number of rating types; Indicates the scoring result.

Images