An automated marketing strategy generation system and method based on intelligent analysis

Through intelligent analysis and automated marketing strategy generation system, the problems of advertising resource allocation and unpopular product promotion in e-commerce platforms are solved, and more efficient advertising delivery and product sales balance is achieved.

CN118863971BActive Publication Date: 2024-11-29JIANGSU BANMA SOFTWARE TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411323780.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-23
Publication Date
2024-11-29
Estimated Expiration
2044-09-23

AI Technical Summary

Technical Problem

The existing e-commerce platforms rely on experience and intuition in the formulation of marketing and promotion strategies, lack data-driven intelligent optimization, resulting in inefficient allocation of advertising resources and the inability to accurately identify advertising needs of different products, resulting in low advertising input-output ratio and unpopular products. The promotion effect is poor and the overall product sales structure cannot be balanced.

Method used

Adopt an automated marketing strategy generation system based on intelligent analysis to screen popular and unpopular products through historical sales data and advertising marketing data, use matching regression models and genetic algorithms to generate target marketing strategies, distinguish advertising-dependent and independent products, and optimize advertising budgets Deliver and increase the exposure and sales of unpopular products through a reasonable product portfolio.

Benefits of technology

Effectively improve the investment-output ratio of advertising, identify potential unpopular products, optimize the promotion of unpopular products, balance the overall product sales structure, and improve advertising delivery efficiency and sales performance of unpopular products.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118863971B_ABST
    Figure CN118863971B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of computer technology, and discloses an automated marketing strategy generation system and method based on intelligent analysis, comprising: screening out a first product set and a second product set according to historical sales data; subdividing a third product set and a fourth product set from the first product set according to the advertising marketing data of each stable popular product; predicting the matching degree of each persistent unpopular product and advertising-dependent product according to a preconfigured matching degree regression model, and subdividing a fifth product set from the second product set according to the matching degree; generating a target marketing strategy based on the third product set and the fifth product set and using a preconfigured genetic algorithm; the present invention can improve the exposure and sales performance of unpopular products through a reasonable product combination strategy.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and more specifically, to an automated marketing strategy generation system and method based on intelligent analysis. Background Art

[0002] In existing e-commerce platforms, the generation or formulation of marketing and promotion strategies plays a vital role in promoting product sales. Usually, the platform relies on manual analysis of historical product sales data and advertising marketing data to formulate corresponding marketing strategies. However, traditional marketing analysis methods often rely on experience and intuition, lacking data-driven intelligent optimization mechanisms, especially in the allocation of advertising resources and the promotion of unpopular products. In addition, traditional marketing analysis methods are unable to accurately identify the advertising needs of different products, especially when allocating advertising resources. It is difficult to effectively distinguish between advertising-dependent products and advertising-independent products. Furthermore, it is impossible to concentrate advertising budgets on products that really need support, resulting in a low advertising return on investment (ROI).

[0003] In addition, existing marketing strategies mostly use a single discount or promotion method to promote unpopular products. Although this can bring a certain sales increase in the short term, it does not fundamentally solve the promotion problem of unpopular products. It is also unable to identify unpopular products with potential from many unpopular products, and ignores the potential correlation between unpopular products with potential and other popular products. As a result, it is impossible to improve the exposure and sales performance of unpopular products through reasonable product combination strategies. At the same time, it is difficult to ensure the balance of the overall product sales structure within the platform. Summary of the invention

[0004] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides an automatic marketing strategy generation system and method based on intelligent analysis.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] In a first aspect, a system for generating an automated marketing strategy based on intelligent analysis is provided, the system comprising:

[0007] A first screening module is used to obtain the historical sales data of each product sold on the platform, and screen out a first product set and a second product set according to the historical sales data; the first product set includes M stable popular products, and the second product set includes N persistent unpopular products, where M and N are both integers greater than zero;

[0008] The second screening module is used to obtain the advertising and marketing data of each stable popular product, and to subdivide the first product set into a third product set and a fourth product set according to the advertising and marketing data; the advertising and marketing data includes the marketing advertisements of each sales product and the promotion effect data of the marketing advertisements; the third product set includes P advertising-independent products, and the fourth product set includes Q advertising-dependent products, where P and Q are both integers greater than zero;

[0009] A product matching module is used to predict the matching degree of each persistent unpopular product and the advertising-dependent product according to a preconfigured matching degree regression model, and to subdivide the second product set into a fifth product set according to the matching degree; the fifth product set includes K potential unpopular products, where K is an integer greater than zero;

[0010] The strategy generation module is used to generate a target marketing strategy based on the third product set and the fifth product set and using a preconfigured genetic algorithm; the target marketing strategy is an optimal product combination, each of which includes an advertising-independent product and a potential unpopular product associated with it.

[0011] Furthermore, the step of selecting the first product set and the second product set according to the historical sales data includes:

[0012] a1: Retrieve historical sales data and calculate the average sales volume and sales volume variance of each product based on the historical sales data;

[0013] The historical sales data includes the length of the evaluation period, the average sales volume of each sales product during the evaluation period, and the sales volume of each sales product at each time t;

[0014] a2: Substitute the average sales volume and sales volume variance of each product into the pre-built piecewise function used to distinguish product types to obtain the market performance evaluation value of each product;

[0015] The functional expression of the piecewise function for distinguishing product types is as follows:

[0016] ;

[0017] Where: To evaluate the market performance; is the sales volume variance; is the average sales volume during the evaluation period; and , respectively represent the minimum threshold and maximum threshold of the preset sales volume variance; and , respectively represent the minimum threshold and maximum threshold of the preset average sales volume; is a natural constant;

[0018] a3: Setting a market performance evaluation threshold, wherein the market performance evaluation threshold includes market performance evaluation thresholds Td1 and Td2, where Td>Td2>0;

[0019] a4: Compare the market performance evaluation value of each sales product with the market performance evaluation threshold; if ≥Td1, the corresponding sales product is determined to be a stable popular product; if ≤Td2, the corresponding sales product is judged to be a persistent unpopular product;

[0020] a5: Repeat step a4 until every selling product is screened, thereby obtaining a first product set including M stable popular products and a second product set including N persistent unpopular products.

[0021] Further, the step of subdividing the first product set into a third product set and a fourth product set according to the advertising and marketing data includes:

[0022] b1: Obtain the marketing advertisements associated with each stable popular product from the advertising marketing data, and extract the promotion effect data of the marketing advertisements associated with each stable popular product;

[0023] The promotion effect data includes the number of marketing advertisements, the sales volume before each marketing advertisement, the sales volume after each marketing advertisement, the click rate of each marketing advertisement, the conversion rate of each marketing advertisement, and the sales growth rate after each marketing advertisement.

[0024] b2: Substitute the promotion effect data into a pre-built mathematical calculation model for reflecting the degree of advertising impact to obtain the advertising impact value of marketing advertising on stable popular products;

[0025] The mathematical expression of the mathematical calculation model for reflecting the degree of advertising influence is as follows:

[0026] ;

[0027] Where: is the advertising impact value; The number of times marketing advertisements are delivered; is the sales volume before the i-th marketing advertisement is launched; is the sales volume after the i-th marketing advertisement is launched; is the click-through rate of the marketing advertisement launched for the i-th time; is the conversion rate of the i-th marketing advertisement; is the sales growth rate after the i-th marketing advertisement is launched; is the base growth rate;

[0028] b3: Compare the advertising impact value with the preset advertising impact threshold; if the advertising impact value is greater than or equal to the advertising impact threshold, the stable popular product associated with the corresponding marketing advertisement is marked as an advertising-dependent product; if the advertising impact value is less than the advertising impact threshold, the stable popular product associated with the corresponding marketing advertisement is marked as an advertising-independent product;

[0029] b4: Repeat step b3 until each stable popular product is screened, and obtain a third product set including P advertising-independent products and a fourth product set including Q advertising-dependent products.

[0030] Furthermore, the predicting of the matching degree between each persistent unpopular product and the advertising-dependent product includes:

[0031] Obtain matching feature data of each persistent unpopular product and advertising-dependent product; the matching feature data includes audience group data, price features, and a product feature set; wherein the audience group data includes age features, gender features, occupation features, income level features, and educational background features; the product feature set includes U product feature sets, where U is an integer greater than zero;

[0032] Input the matching feature data into the pre-configured matching regression model to obtain the matching degree between the persistent unpopular products and the advertising-dependent products;

[0033] The generation logic of the preconfigured matching regression model is as follows:

[0034] Acquire historical matching training data, and divide the historical matching training data into a matching training set and a matching test set, wherein the historical matching training data includes matching feature data of a plurality of persistent unpopular products and advertising-dependent products and corresponding matching degrees;

[0035] Constructing a regression network, taking the matching feature data in the matching training set as the regression network input, and taking the matching degree in the matching training set as the regression network output, training the regression network, and obtaining an initial regression network;

[0036] The matching test set is used to verify the model of the initial regression network, and the initial regression network with a test error less than or equal to the test error is output as the preconfigured matching regression model.

[0037] Furthermore, the logic for generating the matching degree in the historical matching degree training data is as follows:

[0038] Obtain matching feature data of persistent unpopular products and advertising-dependent products, and extract audience group data, price features, and product function feature sets from the matching feature data;

[0039] Calculate the first similarity between persistent unpopular products and advertising-dependent products based on audience data;

[0040] Calculate the second similarity between persistent unpopular products and advertising-dependent products based on price features;

[0041] Calculate the third similarity between persistent unpopular products and advertising-dependent products based on the product feature set;

[0042] Statistically calculate the first similarity, the second similarity, and the third similarity to obtain the matching degree between the persistent unpopular products and the advertising-dependent products;

[0043] The calculation formula for the statistical calculation is as follows:

[0044] ;

[0045] Where: The matching degree between persistent unpopular products and advertising-dependent products; is the first similarity; is the second similarity; is the third similarity; , and is the weight factor, .

[0046] Furthermore, the step of subdividing the fifth product set from the second product set according to the matching degree includes:

[0047] c1: Extract the matching degree between each persistent unpopular product and the advertising-dependent product;

[0048] c2: Compare the matching degree with the preset matching degree threshold; if the matching degree is greater than or equal to the matching degree threshold, the corresponding persistent unpopular product is determined to be a potential unpopular product; if the matching degree is less than the matching degree threshold, the corresponding persistent unpopular product is determined to be a non-potential unpopular product;

[0049] c3: Repeat step c2 until the matching degrees of all persistent unpopular products and advertising-dependent products are extracted, and a fifth product set containing K potential unpopular products is obtained.

[0050] Furthermore, the generating of the target marketing strategy by using the preconfigured genetic algorithm includes:

[0051] d1: Initialization population: randomly generate an original population, which contains X chromosomes, each chromosome represents an initial product combination, each product combination contains an advertising-independent product and a potential unpopular product, and X is an integer greater than zero;

[0052] d2: Fitness evaluation: Under each chromosome (i.e., each product combination in the original population), obtain the sales change of advertising-independent products within a certain period of time, as well as the sales growth rate of potential unpopular products; input the future sales and future exposure rate into the pre-built fitness function to calculate the fitness of each chromosome;

[0053] The calculation formula of the pre-constructed fitness function is: ; Where: For fitness, is the sales volume change of advertising-independent products, is the sales growth rate of potential unpopular products, and is a weight factor greater than zero, ;

[0054] d3: Selection: Use the roulette wheel method to select two chromosomes with high fitness in the original population as the father and mother;

[0055] d4: Crossover: Perform crossover operation on the father and mother to generate new chromosomes;

[0056] d5: mutation: perform mutation operation on the new chromosomes to obtain Y new chromosomes, combine the Y new chromosomes into a new population, replace the original population with the new population, and return to step d2;

[0057] d6: Repeat steps d2 to d5 until the fitness of the chromosome in the original population or the new population is greater than or equal to the preset fitness threshold, or the number of iterations is greater than or equal to the preset maximum number of iterations threshold, and output the product combination represented by the corresponding chromosome as the optimal product combination.

[0058] In a second aspect, a method for generating an automated marketing strategy based on intelligent analysis is provided, the method comprising:

[0059] Obtain historical sales data of each product sold on the platform, and select a first product set and a second product set according to the historical sales data; the first product set includes M stable popular products, and the second product set includes N persistent unpopular products, where M and N are both integers greater than zero;

[0060] Acquire advertising and marketing data of each stable popular product, and subdivide the first product set into a third product set and a fourth product set according to the advertising and marketing data; the advertising and marketing data includes marketing advertisements of each sold product and promotion effect data of the marketing advertisements; the third product set includes P advertising-independent products, and the fourth product set includes Q advertising-dependent products, where P and Q are both integers greater than zero;

[0061] According to a preconfigured matching regression model, the matching degree of each persistent unpopular product and the advertising-dependent product is predicted, and a fifth product set is subdivided from the second product set according to the matching degree; the fifth product set includes K potential unpopular products, where K is an integer greater than zero;

[0062] A target marketing strategy is generated based on the third product set and the fifth product set and by using a preconfigured genetic algorithm; the target marketing strategy is an optimal product combination, each of which includes an advertising-independent product and a potential unpopular product associated with it.

[0063] According to a third aspect, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, the method for generating an automated marketing strategy based on intelligent analysis as described above is implemented.

[0064] In a fourth aspect, a computer-readable storage medium is provided, wherein a computer program is stored on the computer-readable storage medium, and when the computer program is executed, the method for generating an automated marketing strategy based on intelligent analysis described in any one of the above is implemented.

[0065] Compared with the prior art, the present invention has the following beneficial effects:

[0066] The present application discloses an automated marketing strategy generation system and method based on intelligent analysis, including: selecting a first product set and a second product set according to historical sales data; subdividing a third product set and a fourth product set from the first product set according to the advertising marketing data of each stable popular product; predicting the matching degree of each persistent unpopular product with an advertising-dependent product according to a preconfigured matching degree regression model, and subdividing a fifth product set from the second product set according to the matching degree; generating a target marketing strategy based on the third product set and the fifth product set and using a preconfigured genetic algorithm; based on the above technical features, the present invention can effectively distinguish between advertising-dependent products and advertising-independent products, so that the advertising budget can be concentrated on products that really need support, thereby improving the return on investment (ROI) of advertising; in addition, by identifying unpopular products with potential from a large number of unpopular products, and using a genetic algorithm to market the unpopular products with potential and popular products, the present invention can improve the exposure and sales performance of unpopular products through a reasonable product combination strategy, thereby helping to fundamentally solve the promotion problem of unpopular products and ensure the balance of the overall product sales structure within the platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 A schematic diagram of the module structure of an automatic marketing strategy generation system based on intelligent analysis provided by the present invention;

[0068] Figure 2 A flowchart of an automated marketing strategy generation method based on intelligent analysis provided by the present invention;

[0069] Figure 3 The present invention provides a schematic structural diagram of an electronic device. DETAILED DESCRIPTION

[0070] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0071] Example 1

[0072] See also Figure 1 As shown, this embodiment discloses a system for generating an automated marketing strategy based on intelligent analysis, the system comprising:

[0073] The first screening module 101 is used to obtain the historical sales data of each product sold in the platform, and screen out a first product set and a second product set according to the historical sales data; the first product set includes M stable popular products, and the second product set includes N persistent unpopular products, where M and N are both integers greater than zero;

[0074] It should be noted that stable popular products refer to those products that maintain high sales volume for a long period of time with little fluctuation. On the contrary, those products that have low sales volume for a long period of time and no obvious growth signs are persistent unpopular products.

[0075] In implementation, the step of selecting the first product set and the second product set based on historical sales data includes:

[0076] a1: Retrieve historical sales data and calculate the average sales volume and sales volume variance of each product based on the historical sales data;

[0077] The historical sales data includes the length of the evaluation period, the average sales volume of each sales product during the evaluation period, and the sales volume of each sales product at each time t;

[0078] The formula for calculating the average sales volume of each product is as follows:

[0079] ;

[0080] Where: is the average sales volume during the evaluation period; is the sales volume of the product at each time t;

[0081] The formula for calculating the sales volume variance of each product sold is as follows:

[0082] ;

[0083] Where: is the sales volume variance;

[0084] a2: Substitute the average sales volume and sales volume variance of each product into the pre-built piecewise function used to distinguish product types to obtain the market performance evaluation value of each product;

[0085] The functional expression of the piecewise function for distinguishing product types is as follows:

[0086] ;

[0087] Where: To evaluate the market performance; is the sales volume variance; is the average sales volume during the evaluation period; and , respectively represent the minimum threshold and maximum threshold of the preset sales volume variance; and , respectively represent the minimum threshold and maximum threshold of the preset average sales volume; is a natural constant;

[0088] a3: Setting a market performance evaluation threshold, wherein the market performance evaluation threshold includes market performance evaluation thresholds Td1 and Td2, where Td>Td2>0;

[0089] a4: Compare the market performance evaluation value of each sales product with the market performance evaluation threshold; if ≥Td1, the corresponding sales product is determined to be a stable popular product; if ≤Td2, the corresponding sales product is judged to be a persistent unpopular product;

[0090] a5: Repeat step a4 until all the selling products are screened, and obtain a first product set including M stable popular products and a second product set including N persistent unpopular products;

[0091] It should be noted that: by calculating the market performance evaluation value through piecewise function, the stable popular products and the persistent unpopular products can be clearly screened out according to the market performance evaluation value; further explanation is that when ≥Td1, it means that the sales volume of the corresponding product is high, and the sales status is in a long-term stable fluctuation, so the corresponding product is classified as a stable hot product; on the contrary, when When ≤Td2, it means that the sales volume of the corresponding product is low, and the sales status is in a stable fluctuation for a long time, so the corresponding product is classified as a persistent unpopular product;

[0092] It should also be noted that: when the market performance evaluation value of the sales product is between the market performance evaluation thresholds Td1 and Td2, that is, when Td1> >Td2, it means that the sales status fluctuates greatly, that is, the sales volume of the corresponding sales product is in a non-steady state; when the sales volume of the sales product is in a non-steady state, it is marked as a high-volatility product. The sales performance of high-volatility products is between stable popular products and persistent unpopular products, which is neither unstably popular nor continuously sluggish; their sales are more affected by external factors such as seasonality, promotions and discounts than stable popular products and persistent unpopular products, and sometimes perform well and sometimes perform poorly; due to their nonlinear changes, data mining is more difficult, so they are not taken into account for the time being, and they are not suitable for subsequent analysis; among them, , , and It is set by technical personnel based on actual conditions.

[0093] The second screening module 102 is used to obtain advertising and marketing data of each stable popular product, and to subdivide the first product set into a third product set and a fourth product set according to the advertising and marketing data; the advertising and marketing data includes marketing advertisements of each sales product and promotion effect data of the marketing advertisements; the third product set includes P advertising-independent products, and the fourth product set includes Q advertising-dependent products, where P and Q are both integers greater than zero;

[0094] It should be noted that advertising-independent products refer to stable popular products whose sales volume does not rely on advertising promotion (i.e., advertising investment has little or no effect on the sales volume of stable popular products) and can maintain high sales volume in the market. On the contrary, the sales volume of products is directly related to advertising, that is, those stable popular products that are highly dependent on advertising to maintain high sales volume are advertising-dependent products.

[0095] In implementation, the step of subdividing the first product set into a third product set and a fourth product set according to the advertising marketing data includes:

[0096] b1: Obtain the marketing advertisements associated with each stable popular product from the advertising marketing data, and extract the promotion effect data of the marketing advertisements associated with each stable popular product;

[0097] It should be noted that: each sales product is artificially associated with one or more marketing advertisements; each associated marketing advertisement is used to promote the corresponding sales product; the marketing advertisement associated with each stable popular product (or sales product) is pre-stored in the system database;

[0098] The promotion effect data includes the number of marketing advertisements, the sales volume before each marketing advertisement, the sales volume after each marketing advertisement, the click rate of each marketing advertisement, the conversion rate of each marketing advertisement, and the sales growth rate after each marketing advertisement.

[0099] b2: Substitute the promotion effect data into a pre-built mathematical calculation model for reflecting the degree of advertising impact to obtain the advertising impact value of marketing advertising on stable popular products;

[0100] The mathematical expression of the mathematical calculation model for reflecting the degree of advertising influence is as follows:

[0101] ;

[0102] Where: is the advertising impact value; The number of times marketing advertisements are delivered; is the sales volume before the i-th marketing advertisement is launched; is the sales volume after the i-th marketing advertisement is launched; is the click-through rate of the marketing advertisement launched for the i-th time; is the conversion rate of the i-th marketing advertisement; is the sales growth rate after the i-th marketing advertisement is launched; is the benchmark growth rate; the benchmark growth rate is set by technical personnel based on actual conditions;

[0103] b3: Compare the advertising impact value with the preset advertising impact threshold; if the advertising impact value is greater than or equal to the advertising impact threshold, the stable popular product associated with the corresponding marketing advertisement is marked as an advertising-dependent product; if the advertising impact value is less than the advertising impact threshold, the stable popular product associated with the corresponding marketing advertisement is marked as an advertising-independent product;

[0104] b4: Repeat step b3 until all stable popular products are screened, and obtain a third product set including P advertising-independent products and a fourth product set including Q advertising-dependent products;

[0105] It can be understood that: the larger the advertising impact value, the more the sales of the corresponding stable popular product depends on advertising promotion. On the contrary, the smaller the advertising impact value, the less the sales of the corresponding stable popular product depends on advertising promotion.

[0106] It should be noted that: for advertising-independent products, the impact of advertising promotion is relatively small, or the sales volume of such products is stable, and there is no need to invest too much advertising capital. For advertising-dependent products, advertising investment should continue to be maintained or increased; further, by screening out products that are conducive to concentrating advertising budgets on products that really need support, the advertising return on investment (ROI) can be improved.

[0107] The product matching module 103 is used to predict the matching degree of each persistent unpopular product and the advertising-dependent product according to a preconfigured matching degree regression model, and to subdivide the second product set into a fifth product set according to the matching degree; the fifth product set includes K potential unpopular products, where K is an integer greater than zero;

[0108] Among them, potential unpopular products refer to those products that may have low current sales due to inaccurate advertising positioning or insufficient advertising promotion, but have the potential for market sales growth;

[0109] In implementation, the predicting of the matching degree between each persistent unpopular product and the advertising-dependent product includes:

[0110] Obtain matching feature data of each persistent unpopular product and advertising-dependent product; the matching feature data includes audience group data, price features, and a product feature set; wherein the audience group data includes but is not limited to age features, gender features, occupational features, income level features, and educational background features, etc.; the product feature set includes U product feature sets, where U is an integer greater than zero;

[0111] Input the matching feature data into the pre-configured matching regression model to obtain the matching degree between the persistent unpopular products and the advertising-dependent products;

[0112] The generation logic of the preconfigured matching regression model is as follows:

[0113] Acquire historical matching training data, and divide the historical matching training data into a matching training set and a matching test set, wherein the historical matching training data includes matching feature data of a plurality of persistent unpopular products and advertising-dependent products and corresponding matching degrees;

[0114] The matching feature data in the historical matching degree training data is actually collected by the technicians according to the historical data records;

[0115] The logic for generating the matching degree in the historical matching degree training data is as follows:

[0116] Obtain matching feature data of persistent unpopular products and advertising-dependent products, and extract audience group data, price features, and product function feature sets from the matching feature data;

[0117] Calculate the first similarity between persistent unpopular products and advertising-dependent products based on audience data;

[0118] The calculation formula of the first similarity is as follows:

[0119] ;

[0120] Where: is the first similarity; Represents the total number of features in the audience data; Represents the v-th vector feature in the audience data of a persistently unpopular product; The v-th vector-like feature in the audience data of an ad-dependent product; represents the norm;

[0121] Calculate the second similarity between persistent unpopular products and advertising-dependent products based on price features;

[0122] The calculation formula of the second similarity is as follows:

[0123] ;

[0124] Where: is the second similarity; Indicates the price characteristics of persistent unpopular products; Indicates the price characteristics of advertising-dependent products; represents the maximum value function;

[0125] Calculate the third similarity between persistent unpopular products and advertising-dependent products based on the product feature set;

[0126] The calculation formula of the third similarity is as follows:

[0127] ;

[0128] Where: is the third similarity; The intersection of the product feature sets of persistent unpopular products and advertising-dependent products; The union of the product feature sets of persistent unpopular products and advertising-dependent products;

[0129] Statistically calculate the first similarity, the second similarity, and the third similarity to obtain the matching degree between the persistent unpopular products and the advertising-dependent products;

[0130] The calculation formula for the statistical calculation is as follows:

[0131] ;

[0132] Where: The matching degree between persistent unpopular products and advertising-dependent products; is the first similarity; is the second similarity; is the third similarity; , and is the weight factor, ;

[0133] Constructing a regression network, taking the matching feature data in the matching training set as the regression network input, and taking the matching degree in the matching training set as the regression network output, training the regression network, and obtaining an initial regression network;

[0134] The initial regression network is model verified using the matching test set, and the initial regression network with a test error less than or equal to the test error is output as the preconfigured matching regression model;

[0135] It should be noted that: the regression network is specifically one of the regression algorithm models such as decision tree, random forest, support vector machine, polynomial regression, convolutional neural network or recurrent neural network;

[0136] In implementation, the step of subdividing the fifth product set from the second product set according to the matching degree includes:

[0137] c1: Extract the matching degree between each persistent unpopular product and the advertising-dependent product;

[0138] c2: Compare the matching degree with the preset matching degree threshold; if the matching degree is greater than or equal to the matching degree threshold, the corresponding persistent unpopular product is determined to be a potential unpopular product; if the matching degree is less than the matching degree threshold, the corresponding persistent unpopular product is determined to be a non-potential unpopular product;

[0139] c3: Repeat step c2 until the matching degrees of all persistent unpopular products and advertising-dependent products are extracted, and a fifth product set containing K potential unpopular products is obtained.

[0140] The strategy generation module 104 is used to generate a target marketing strategy based on the third product set and the fifth product set and using a preconfigured genetic algorithm; the target marketing strategy is an optimal product combination, each of which includes an advertising-independent product and a potential unpopular product associated with it;

[0141] In implementation, the method of generating a target marketing strategy using a preconfigured genetic algorithm includes:

[0142] d1: Initialization population: randomly generate an original population, which contains X chromosomes, each chromosome represents an initial product combination, each product combination contains an advertising-independent product and a potential unpopular product, and X is an integer greater than zero;

[0143] d2: Fitness evaluation: Under each chromosome (i.e., each product combination in the original population), obtain the sales change of advertising-independent products within a certain period of time, as well as the sales growth rate of potential unpopular products; input the future sales and future exposure rate into the pre-built fitness function to calculate the fitness of each chromosome;

[0144] The calculation formula of the pre-constructed fitness function is: ; Where: For fitness, is the sales volume change of advertising-independent products, is the sales growth rate of potential unpopular products, and is a weight factor greater than zero, ;

[0145] It should be noted that the sales volume change is obtained by subtracting the sales volume of the advertising-independent product after the combination from the sales volume of the advertising-independent product before the combination. The fitness function reveals the optimization goal of the best product combination, that is, to increase the sales volume of the potential unpopular product as much as possible without generating negative impact or generating a small amount of negative impact. Among them, the negative impact refers to the negative growth rate of the sales volume of the advertising-independent product.

[0146] d3: Selection: Use the roulette wheel method to select two chromosomes with high fitness in the original population as the father and mother;

[0147] d4: Crossover: Perform crossover operation on the father and mother to generate new chromosomes;

[0148] The crossover operation on the paternal parent and the maternal parent is implemented based on a crossover operation, and the crossover operation is not limited to one of a single-point crossover, a uniform crossover, or a sequential crossover, etc.;

[0149] d5: mutation: perform mutation operation on the new chromosomes to obtain Y new chromosomes, combine the Y new chromosomes into a new population, replace the original population with the new population, and return to step d2;

[0150] The mutation operation on the new chromosome is implemented by uniform mutation or Gaussian mutation;

[0151] d6: Repeat steps d2 to d5 until the fitness of the chromosome in the original population or the new population is greater than or equal to the preset fitness threshold, or the number of iterations is greater than or equal to the preset maximum number of iterations threshold, and output the product combination represented by the corresponding chromosome as the optimal product combination;

[0152] For example: Assume that the maximum number of iterations is 100 times, and after each iteration, record the chromosome with the highest fitness in the current population and its fitness value; if it is found that the fitness value has not changed significantly in a certain generation, it is considered that the convergence condition is met, the iteration is stopped, and the product combination represented by the corresponding chromosome is output as the best product combination;

[0153] It can be understood that: by effectively distinguishing between advertising-dependent products and advertising-independent products, the present invention is conducive to concentrating advertising budgets on products that really need support, thereby improving the advertising return on investment (ROI); in addition, by identifying unpopular products with potential from many unpopular products, and using genetic algorithms to market potential unpopular products with popular products, the present invention can improve the exposure and sales performance of unpopular products through reasonable product combination strategies, which is conducive to fundamentally solving the promotion problem of unpopular products and ensuring the balance of the overall product sales structure within the platform.

[0154] Example 2

[0155] See also Figure 2 As shown, based on the same inventive concept, this embodiment discloses a method for generating an automated marketing strategy based on intelligent analysis. For details not provided in this embodiment, please refer to the description of the relevant parts in Embodiment 1. The method includes:

[0156] S201: Obtain historical sales data of each product sold on the platform, and select a first product set and a second product set according to the historical sales data; the first product set includes M stable popular products, and the second product set includes N persistent unpopular products, where M and N are both integers greater than zero;

[0157] S202: Acquire advertising marketing data for each stable popular product, and subdivide the first product set into a third product set and a fourth product set according to the advertising marketing data; the advertising marketing data includes marketing advertisements for each sold product and promotion effect data of the marketing advertisements; the third product set includes P advertising-independent products, and the fourth product set includes Q advertising-dependent products, where P and Q are both integers greater than zero;

[0158] S203: predicting the matching degree of each persistent unpopular product and the advertising-dependent product according to a preconfigured matching degree regression model, and subdividing a fifth product set from the second product set according to the matching degree; the fifth product set includes K potential unpopular products, where K is an integer greater than zero;

[0159] S204: Generate a target marketing strategy based on the third product set and the fifth product set and using a preconfigured genetic algorithm; the target marketing strategy is an optimal product combination, each of which includes an advertising-independent product and a potential unpopular product associated with it.

[0160] Example 3

[0161] See also Figure 3 As shown, this embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, it implements any one of the methods provided by the above methods to generate an automated marketing strategy based on intelligent analysis.

[0162] Since the electronic device introduced in the content of this embodiment is an electronic device used to implement an automatic marketing strategy generation method based on intelligent analysis in the embodiment of this application, based on the automatic marketing strategy generation method based on intelligent analysis introduced in the embodiment of this application, the technical personnel of this field can understand the specific implementation of the electronic device of this embodiment and its various variations, so how the electronic device implements the method in the embodiment of this application is not introduced in detail here. As long as the technical personnel of this field implement the electronic device used in the automatic marketing strategy generation method based on intelligent analysis in the embodiment of this application, it belongs to the scope of protection of this application.

[0163] Example 4

[0164] This embodiment discloses a computer-readable storage medium, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements any one of the methods provided by the above methods to generate an automated marketing strategy based on intelligent analysis.

[0165] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters, weights and thresholds in the formula are set by technicians in this field according to actual conditions.

[0166] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server or data center to another website, computer, server or data center via a wired network or a wireless network. The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media sets. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD) or a semiconductor medium. The semiconductor medium may be a solid-state hard disk.

[0167] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0168] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0169] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only one, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0170] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0171] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0172] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0173] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An automated marketing strategy generation system based on intelligent analysis, characterized in that: The system comprises: A first screening module is used to obtain the historical sales data of each product sold on the platform, and screen out a first product set and a second product set according to the historical sales data; the first product set includes M stable popular products, and the second product set includes N persistent unpopular products, where M and N are both integers greater than zero; The second screening module is used to obtain the advertising and marketing data of each stable popular product, and to subdivide the first product set into a third product set and a fourth product set according to the advertising and marketing data; the advertising and marketing data includes the marketing advertisements of each sales product and the promotion effect data of the marketing advertisements; the third product set includes P advertising-independent products, and the fourth product set includes Q advertising-dependent products, where P and Q are both integers greater than zero; A product matching module is used to predict the matching degree of each persistent unpopular product and the advertising-dependent product according to a preconfigured matching degree regression model, and to subdivide the second product set into a fifth product set according to the matching degree; the fifth product set includes K potential unpopular products, where K is an integer greater than zero; The predicting of the matching degree between each persistent unpopular product and the advertising-dependent product includes: Obtain matching feature data of each persistent unpopular product and advertising-dependent product; the matching feature data includes audience group data, price features, and a product feature set; wherein the audience group data includes age features, gender features, occupation features, income level features, and educational background features; the product feature set includes U product feature sets, where U is an integer greater than zero; Input the matching feature data into the pre-configured matching regression model to obtain the matching degree between the persistent unpopular products and the advertising-dependent products; The generation logic of the preconfigured matching regression model is as follows: Acquire historical matching training data, and divide the historical matching training data into a matching training set and a matching test set, wherein the historical matching training data includes matching feature data of a plurality of persistent unpopular products and advertising-dependent products and corresponding matching degrees; Constructing a regression network, taking the matching feature data in the matching training set as the regression network input, and taking the matching degree in the matching training set as the regression network output, training the regression network, and obtaining an initial regression network; The initial regression network is model verified using the matching test set, and the initial regression network with a test error less than or equal to the test error is output as the preconfigured matching regression model; The logic for generating the matching degree in the historical matching degree training data is as follows: Obtain matching feature data of persistent unpopular products and advertising-dependent products, and extract audience group data, price features, and product function feature sets from the matching feature data; Calculate the first similarity between persistent unpopular products and advertising-dependent products based on audience data; Calculate the second similarity between persistent unpopular products and advertising-dependent products based on price features; Calculate the third similarity between persistent unpopular products and advertising-dependent products based on the product feature set; Statistically calculate the first similarity, the second similarity, and the third similarity to obtain the matching degree between the persistent unpopular products and the advertising-dependent products; The calculation formula for the statistical calculation is as follows: ; Where: The matching degree between persistent unpopular products and advertising-dependent products; is the first similarity; is the second similarity; is the third similarity; , and is the weight factor, ; The step of subdividing the fifth product set from the second product set according to the matching degree includes: c1: Extract the matching degree between each persistent unpopular product and the advertising-dependent product; c2: Compare the matching degree with the preset matching degree threshold; if the matching degree is greater than or equal to the matching degree threshold, the corresponding persistent unpopular product is determined to be a potential unpopular product; if the matching degree is less than the matching degree threshold, the corresponding persistent unpopular product is determined to be a non-potential unpopular product; c3: Repeat step c2 until the matching degree of all persistent unpopular products and advertising-dependent products is extracted, and a fifth product set containing K potential unpopular products is obtained; A strategy generation module is used to generate a target marketing strategy based on the third product set and the fifth product set and using a preconfigured genetic algorithm; the target marketing strategy is an optimal product combination, each of which includes an advertising-independent product and a potential unpopular product associated with it; Wherein, the method of generating a target marketing strategy by using a preconfigured genetic algorithm includes: d1: Initialization population: randomly generate an original population, which contains X chromosomes, each chromosome represents an initial product combination, each product combination contains an advertising-independent product and a potential unpopular product, and X is an integer greater than zero; d2: Fitness evaluation: Under each chromosome, obtain the sales change of advertising-independent products within a certain period of time, as well as the sales growth rate of potential unpopular products; input future sales and future exposure rates into the pre-built fitness function to calculate the fitness of each chromosome; The calculation formula of the pre-constructed fitness function is: ; Where: For fitness, is the sales volume change of advertising-independent products, is the sales growth rate of potential unpopular products, and is a weight factor greater than zero, ; d3: Selection: Use the roulette wheel method to select two chromosomes with high fitness in the original population as the father and mother; d4: Crossover: Perform crossover operation on the father and mother to generate new chromosomes; d5: mutation: perform mutation operation on the new chromosomes to obtain Y new chromosomes, combine the Y new chromosomes into a new population, replace the original population with the new population, and return to step d2; d6: Repeat steps d2 to d5 until the fitness of the chromosome in the original population or the new population is greater than or equal to the preset fitness threshold, or the number of iterations is greater than or equal to the preset maximum number of iterations threshold, and output the product combination represented by the corresponding chromosome as the optimal product combination.

2. The automated marketing strategy generation system based on intelligent analysis according to claim 1, characterized in that: The step of selecting the first product set and the second product set according to the historical sales data includes: a1: Retrieve historical sales data and calculate the average sales volume and sales volume variance of each product based on the historical sales data; The historical sales data includes the length of the evaluation period, the average sales volume of each sales product during the evaluation period, and the sales volume of each sales product at each time t; a2: Substitute the average sales volume and sales volume variance of each product into the pre-built piecewise function used to distinguish product types to obtain the market performance evaluation value of each product; The functional expression of the piecewise function for distinguishing product types is as follows: ; Where: To evaluate the market performance; is the sales volume variance; is the average sales volume during the evaluation period; and , respectively represent the minimum threshold and maximum threshold of the preset sales volume variance; and , respectively represent the minimum threshold and maximum threshold of the preset average sales volume; is a natural constant; a3: Setting a market performance evaluation threshold, wherein the market performance evaluation threshold includes market performance evaluation thresholds Td1 and Td2, where Td>Td2>0; a4: Compare the market performance evaluation value of each sales product with the market performance evaluation threshold; if ≥Td1, the corresponding sales product is determined to be a stable popular product; if ≤Td2, the corresponding sales product is judged to be a persistent unpopular product; a5: Repeat step a4 until every selling product is screened, thereby obtaining a first product set including M stable popular products and a second product set including N persistent unpopular products.

3. The automated marketing strategy generation system based on intelligent analysis according to claim 2 is characterized in that: The step of subdividing the first product set into a third product set and a fourth product set according to the advertising and marketing data includes: b1: Obtain the marketing advertisements associated with each stable popular product from the advertising marketing data, and extract the promotion effect data of the marketing advertisements associated with each stable popular product; The promotion effect data includes the number of marketing advertisements, the sales volume before each marketing advertisement, the sales volume after each marketing advertisement, the click rate of each marketing advertisement, the conversion rate of each marketing advertisement, and the sales growth rate after each marketing advertisement. b2: Substitute the promotion effect data into a pre-built mathematical calculation model for reflecting the degree of advertising impact to obtain the advertising impact value of marketing advertising on stable popular products; The mathematical expression of the mathematical calculation model for reflecting the degree of advertising influence is as follows: ; Where: is the advertising impact value; The number of times marketing advertisements are delivered; is the sales volume before the i-th marketing advertisement is launched; is the sales volume after the i-th marketing advertisement is launched; is the click-through rate of the marketing advertisement launched for the i-th time; is the conversion rate of the i-th marketing advertisement; is the sales growth rate after the i-th marketing advertisement is launched; is the base growth rate; b3: Compare the advertising impact value with the preset advertising impact threshold; if the advertising impact value is greater than or equal to the advertising impact threshold, the stable popular product associated with the corresponding marketing advertisement is marked as an advertising-dependent product; if the advertising impact value is less than the advertising impact threshold, the stable popular product associated with the corresponding marketing advertisement is marked as an advertising-independent product; b4: Repeat step b3 until each stable popular product is screened, and obtain a third product set including P advertising-independent products and a fourth product set including Q advertising-dependent products.

4. A method for generating an automated marketing strategy based on intelligent analysis, characterized in that: The method comprises: Obtain historical sales data of each product sold on the platform, and select a first product set and a second product set according to the historical sales data; the first product set includes M stable popular products, and the second product set includes N persistent unpopular products, where M and N are both integers greater than zero; Acquire advertising and marketing data of each stable popular product, and subdivide the first product set into a third product set and a fourth product set according to the advertising and marketing data; the advertising and marketing data includes marketing advertisements of each sold product and promotion effect data of the marketing advertisements; the third product set includes P advertising-independent products, and the fourth product set includes Q advertising-dependent products, where P and Q are both integers greater than zero; According to a preconfigured matching regression model, the matching degree of each persistent unpopular product and the advertising-dependent product is predicted, and a fifth product set is subdivided from the second product set according to the matching degree; the fifth product set includes K potential unpopular products, where K is an integer greater than zero; The predicting of the matching degree between each persistent unpopular product and the advertising-dependent product includes: Obtain matching feature data of each persistent unpopular product and advertising-dependent product; the matching feature data includes audience group data, price features, and a product feature set; wherein the audience group data includes age features, gender features, occupation features, income level features, and educational background features; the product feature set includes U product feature sets, where U is an integer greater than zero; Input the matching feature data into the pre-configured matching regression model to obtain the matching degree between the persistent unpopular products and the advertising-dependent products; The generation logic of the preconfigured matching regression model is as follows: Acquire historical matching training data, and divide the historical matching training data into a matching training set and a matching test set, wherein the historical matching training data includes matching feature data of a plurality of persistent unpopular products and advertising-dependent products and corresponding matching degrees; Constructing a regression network, taking the matching feature data in the matching training set as the regression network input, and taking the matching degree in the matching training set as the regression network output, training the regression network, and obtaining an initial regression network; The initial regression network is model verified using the matching test set, and the initial regression network with a test error less than or equal to the test error is output as the preconfigured matching regression model; The logic for generating the matching degree in the historical matching degree training data is as follows: Obtain matching feature data of persistent unpopular products and advertising-dependent products, and extract audience group data, price features, and product function feature sets from the matching feature data; Calculate the first similarity between persistent unpopular products and advertising-dependent products based on audience data; Calculate the second similarity between persistent unpopular products and advertising-dependent products based on price features; Calculate the third similarity between persistent unpopular products and advertising-dependent products based on the product feature set; Statistically calculate the first similarity, the second similarity, and the third similarity to obtain the matching degree between the persistent unpopular products and the advertising-dependent products; The calculation formula for the statistical calculation is as follows: ; Where: The matching degree between persistent unpopular products and advertising-dependent products; is the first similarity; is the second similarity; is the third similarity; , and is the weight factor, ; The step of subdividing the fifth product set from the second product set according to the matching degree includes: c1: Extract the matching degree between each persistent unpopular product and the advertising-dependent product; c2: Compare the matching degree with the preset matching degree threshold; if the matching degree is greater than or equal to the matching degree threshold, the corresponding persistent unpopular product is determined to be a potential unpopular product; if the matching degree is less than the matching degree threshold, the corresponding persistent unpopular product is determined to be a non-potential unpopular product; c3: Repeat step c2 until the matching degree of all persistent unpopular products and advertising-dependent products is extracted, and a fifth product set containing K potential unpopular products is obtained; Generate a target marketing strategy based on the third product set and the fifth product set and using a preconfigured genetic algorithm; the target marketing strategy is an optimal product combination, each of which includes an advertising-independent product and a potential unpopular product associated with it; Wherein, the method of generating a target marketing strategy by using a preconfigured genetic algorithm includes: d1: Initialization population: randomly generate an original population, which contains X chromosomes, each chromosome represents an initial product combination, each product combination contains an advertising-independent product and a potential unpopular product, and X is an integer greater than zero; d2: Fitness evaluation: Under each chromosome, obtain the sales change of advertising-independent products within a certain period of time, as well as the sales growth rate of potential unpopular products; input future sales and future exposure rates into the pre-built fitness function to calculate the fitness of each chromosome; The calculation formula of the pre-constructed fitness function is: ; Where: For fitness, is the sales volume change of advertising-independent products, is the sales growth rate of potential unpopular products, and is a weight factor greater than zero, ; d3: Selection: Use the roulette wheel method to select two chromosomes with high fitness in the original population as the father and mother; d4: Crossover: Perform crossover operation on the father and mother to generate new chromosomes; d5: mutation: perform mutation operation on the new chromosomes to obtain Y new chromosomes, combine the Y new chromosomes into a new population, replace the original population with the new population, and return to step d2; d6: Repeat steps d2 to d5 until the fitness of the chromosome in the original population or the new population is greater than or equal to the preset fitness threshold, or the number of iterations is greater than or equal to the preset maximum number of iterations threshold, and output the product combination represented by the corresponding chromosome as the optimal product combination.

5. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the method for generating an automated marketing strategy based on intelligent analysis as described in claim 4 is implemented.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed, the method for generating an automated marketing strategy based on intelligent analysis as described in claim 4 is implemented.

Citation Information

Patent Citations

  • Commodity marketing management and control method and system based on big data

    CN113781186A

  • Product prediction sales volume-based marketing strategy making method and device

    CN117710011A