A method and system for customer screening and advertisement recommendation based on data analysis

Abstract

The present invention discloses a customer screening and advertisement recommendation method and system based on data analysis, which relates to the field of commercial recommendation. By collecting data of each customer according to various influencing factors, the present invention can comprehensively collect customer information from multiple aspects, making subsequent analysis more comprehensive and accurate; classifying the collected data one by one, and then overall classifying the classified category data, so as to determine the category of customers under the above-mentioned influencing factors, and then sorting the customers according to their categories, and screening the customers conveniently according to the sorting results; since the advertisement distribution is often targeted at customers, the selection of advertisement types in the present invention is based on the advertisement types already distributed to the screened customers themselves, so that the advertisement method has stronger pertinence and higher efficiency.

Description

Technical Field

[0001] The present invention belongs to the field of commercial recommendation. Specifically, it particularly relates to a method and system for customer screening and advertisement recommendation based on data analysis. Background Art

[0002] Customer screening and advertisement recommendation is a marketing means based on data analysis, aiming to help enterprises more effectively identify and attract potential customers, while improving the efficiency and return rate of advertisement placement; the customer screening process involves identifying and classifying those people who are most likely to have a demand for products or services and are most likely to become long-term customers; this process generally includes the following steps: market research, data collection, data analysis, customer segmentation, and strategy formulation; the advertisement recommendation system uses the results of customer screening to provide a personalized advertisement experience; these systems generally work as follows: user profiling, advertisement content matching, real-time placement, and effect tracking; the purpose of the customer screening and advertisement recommendation system is to improve the pertinence and efficiency of marketing activities, reduce ineffective advertisement placement, and at the same time enhance the user's advertisement experience; in this way, enterprises can more effectively attract and retain customers in a highly competitive market.

[0003] Chinese Patent CN116385076A discloses an advertisement recommendation method, device, terminal device, and computer-readable storage medium, which acquires advertisement click and exposure behavior data of users who need advertisement recommendation; recalls advertisements according to a preset advertisement recall logic rule to form an initial advertisement set; inputs the advertisement click and exposure behavior data and the initial advertisement set into a pre-trained LightGBM model to calculate the click-through rate scores of each advertisement in the initial advertisement set; selects a preset number of advertisements from the initial advertisement set based on each click-through rate score and recommends them to the user; this method can accurately calculate the click-through rate scores of advertisements based on the advertisement click and exposure behavior data of users and by using the LightGBM model, and thus can provide the accuracy rate of recommended advertisements, greatly reducing the time for users to screen advertisements; in addition, by using this method, differential advertisement recommendations can be made for different users to meet the needs of different customers.

[0004] In the process of existing enterprises screening customers, it usually requires screening personnel to have professional data analysis skills, which may require enterprises to train employees or hire talents with relevant skills, increasing the screening cost of enterprises. Moreover, screening enterprises subjectively by people may be affected by subjective emotions, resulting in inaccurate screening results. Summary of the Invention

[0005] In view of the problems in the related art, the present invention proposes a method and system for customer screening and advertisement recommendation based on data analysis to overcome the above-mentioned technical problems existing in the related art.

[0006] To solve the above technical problems, the present invention is realized through the following technical solutions:

[0007] The present invention relates to a method for customer screening and advertisement recommendation based on data analysis, comprising the following steps:

[0008] S1. Set a customer priority screening factor set and a customer elimination screening factor set; set a set of customers to be screened; collect data on customer priority screening factors of the customers to be screened according to the set of customers to be screened and the customer priority screening factor set; then collect data on customer elimination screening factors of the customers to be screened according to the set of customers to be screened and the customer elimination screening factor set; classify the data on customer priority screening factors of the customers to be screened and the data on customer elimination screening factors of the customers to be screened, and finally obtain a first weighted data category matrix of customers to be screened and a second weighted data category matrix of customers to be screened;

[0009] S2. Construct a first cat population and a second cat population to classify and optimize the first weighted data category matrix of customers to be screened and the second weighted data category matrix of customers to be screened respectively, and obtain an optimal category set of customer priority screening factor data and an optimal category set of customer elimination screening factor data;

[0010] S3. Set a priority screening classification threshold and an elimination screening classification threshold; rank the optimal category sets of customer priority screening factor data and customer elimination screening factor data, and screen and eliminate the customers in the ranked optimal category sets of customer priority screening factor data and customer elimination screening factor data according to the priority screening classification threshold and the elimination screening classification threshold respectively, so as to obtain a final customer set;

[0011] S4. Collect a corresponding final customer advertisement type matrix according to the final customer set; set a set of advertisements to be screened and collect the types of each advertisement to be screened in the set of advertisements to be screened, so as to obtain a set of advertisement types to be screened, compare the set of advertisement types to be screened with the final customer advertisement type matrix, and obtain an advertisement recommendation set; finally, recommend the advertisement recommendation set to enterprise decision-makers;

[0012] The customer priority screening factor set includes, for example, consumption habits, browsing history, social media activities, the needs, preferences, behavior patterns, purchase probability, and consumption ability of the customer's target market. The customer elimination screening factor set includes, for example, poor customer credit records and whether there are any illegal records; collecting data of each customer according to the above influencing factors can comprehensively collect customer information from multiple aspects, making subsequent analysis more comprehensive and accurate; classifying the collected data one by one, and then overall classifying the already classified category data, so as to determine the category of the customer under the above influencing factors, and thus the fit between the customer and the enterprise and the goodness and badness of the customer itself can be separated according to the category, and then sorting the customers according to the category. According to the sorting result, it is convenient to screen the customers; since the advertisement is often targeted at customers, in the solution of the present invention, the selection of the advertisement type is based on the advertisement types that have already been issued to the screened customers themselves, so that the advertisement method has stronger pertinence and higher efficiency.

[0013] Preferably, the S1 includes the following steps:

[0014] S11. Set the customer priority screening factor set a = {a1, a2,..., a j ,..., a a′} and the customer elimination screening factor set a j represents the jth customer priority screening factor, and a′ represents the total number of set customer priority screening factors; represents the jth customer elimination screening factor, represents the total number of set customer priority screening factors; set the customer set to be screened b = {b1, b2,..., b i ,..., b b′}, b i represents the ith customer to be screened, and b′ represents the total number of set customers to be screened;

[0015] S12. According to the customer set to be screened b = {b1, b2,..., b i ,..., b b′} and the customer priority screening factor set a = {a1, a2,..., a j ,..., a a′}, collect the customer priority screening factor data to be screened, and obtain the first customer data matrix to be screened as follows,

[0016]

[0017] wherein, Represents the data of the \(i\)-th customer to be screened in the \(j\)-th customer priority screening factor;

[0018] Then, according to the set of customers to be screened \(b = \{b_1, b_2, \cdots, b_{}\) i , \cdots, b_{} b′}\) and the set of customer elimination screening factors Collect the data of the customer elimination screening factors of the customers to be screened to obtain the second matrix of customers to be screened data as follows,

[0019]

[0020] wherein, Represents the data of the \(i\)-th customer to be screened in the \(j\)-th customer elimination screening factor;

[0021] S13. Set the customer priority screening level set \(c = \{c_1, c_2, \cdots, c_{}\) i , \cdots, c_{} c′}\), \(c_{}\) i Represents the \(i\)-th customer priority screening level, and \(c'\) represents the total number of set customer priority screening levels; According to the customer priority screening factor set \(a = \{a_1, a_2, \cdots, a_{}\) j , \cdots, a_{} a′}\) and the customer priority screening level set \(c = \{c_1, c_2, \cdots, c_{}\) i , \cdots, c_{} c′ Set the customer priority screening level data interval matrix as follows

[0022]

[0023] wherein, and respectively represent the upper and lower limits of the data interval of the \(i\)-th customer priority screening level on the \(j\)-th customer priority screening factor;

[0024] Set the customer elimination screening level set Represents the \(i\)-th customer elimination screening level, and \(d\) represents the total number of set customer elimination screening levels; According to the customer elimination screening factor set and the customer elimination screening level set Set the customer elimination screening level data interval matrix \(d'\); as follows,

[0025]

[0026] wherein, and respectively represent the upper and lower limits of the data interval of the \(i\)-th customer elimination screening level on the \(j\)-th customer elimination screening factor;

[0027] S14. According to the customer priority screening level data interval matrix classify the first customer data matrix to be screened to obtain the first customer data category matrix to be screened as follows

[0028]

[0029] wherein represents the data category of the i-th customer to be screened in the j-th customer priority screening factor;

[0030] Then, according to the customer elimination screening level data interval matrix d′, classify the second customer data matrix to be screened to obtain the second customer data category matrix to be screened as follows

[0031]

[0032] wherein represents the data category of the i-th customer to be screened in the j-th customer elimination screening factor;

[0033] S15. Set the customer priority screening factor weight set f = {f1, f2,... f j ,..., f a′} and the customer elimination screening factor weight set f j represents the weight value of the j-th customer priority screening factor; f i ′ represents the weight value of the i-th customer elimination screening factor;

[0034] According to the customer priority screening factor weight set f = {f1, f2,... f j ,..., f a′} and the first customer data category matrix to be screened obtain the first customer weighted data category matrix Then, according to the customer elimination screening factor weight set and the second customer data category matrix to be screened obtain the second customer weighted data category matrix The first customer weighted data category matrix and the second customer weighted data category matrix are respectively as follows

[0035]

[0036] wherein represents the weighted data category of the i-th customer to be screened in the j-th customer priority screening factor; represents the weighted data category of the i-th customer to be screened in the j-th customer elimination screening factor;

[0037] By setting the customer priority screening factor set and the customer elimination screening factor set, it provides a collection direction for the subsequent collection of customer information, making the collection of customer information more targeted and comprehensive; then, by collecting the first customer data matrix to be screened and the second customer data matrix to be screened, it provides data support for the subsequent classification of customers; by classifying each data in the first customer data matrix to be screened and the second customer data matrix to be screened, it makes the subsequent overall classification of customers more convenient and efficient.

[0038] Preferably, the first weighted data category matrix of customers to be screened in S15 and the second weighted data category matrix of customers to be screened The calculation formulas for the data in are as follows:

[0039]

[0040] Since the importance degrees of each influencing factor are different, through the weighting operation, the accuracy of the data is improved, thereby also improving the accuracy of the subsequent classification of customers.

[0041] Preferably, S2 includes the following steps:

[0042] S21. Construct the first cat population and the second cat population, set the scale of the first cat population as e, the scale of the second cat population as g′, and the first cat population and the second cat population are respectively represented as e′ = {e1′, e′2,..., e i ′,..., e e ′} and e i ′ represents the i-th cat in the first cat population, represents the i-th cat in the second cat population; set the first maximum number of iterations as The second maximum number of iterations is Set the number of position space dimensions of the cats in the first cat population as The number of position space dimensions of the cats in the second cat population is h; set the grouping rate of the first cat population as m, and the grouping rate of the second cat population as m′;

[0043] S22. According to the number of position space dimensions of the cats in the first cat population Set the total category set of customer priority screening factor data \(g = \{g_1, g_2, \ldots, g i , \ldots, g e～ \}\), where \(g i represents the \(i\)-th total category of customer priority screening factor data; then, according to the number of position space dimensions \(h\) of the cats in the second cat population, set the total category set of customer elimination screening factor data \(h'=\{h_1', h_2', \ldots, h i ', \ldots, h h '\}\), where \(h i ' represents the \(i\)-th total category of customer elimination screening factor data;

[0044] S23. Randomly classify each data in the first weighted data category matrix of the to-be-screened customers into the total category set of customer priority screening factor data to obtain the initial customer priority screening factor classification data set Then, randomly classify each data in the second weighted data category matrix of the to-be-screened customers into the total category set \(h'=\{h_1', h_2', \ldots, h i ', \ldots, h h '\}\) of customer elimination screening factor data to obtain the initial customer elimination screening factor classification data set represents the \(i\)-th initial customer priority screening factor classification data, represents the \(i\)-th initial customer elimination screening factor classification data;

[0045] S24. Calculate the classification centers of each classification data in the initial customer priority screening factor classification data set and the initial customer elimination screening factor classification data set respectively to obtain the first initial classification center set and the second initial classification center set \(l'=\{l_1', l_2', \ldots, l i ', \ldots, l h '\}\); \(l i represents the classification center of , and \(l i ' represents the classification center of ;

[0046] S25. Take the first initial classification center set and the second initial classification center set \(l'=\{l_1', l_2', \ldots, l i ', \ldots, l h '\}\) as the positions of the first cat \(e_1'\) in the first cat population and the position of the first cat in the second cat population respectively;

[0047] S26. Repeat S23, S24, and S25 to finally obtain the initial position set of the first cat population

[0048] and the initial position set of the second cat population represents the initial position of the i-th cat in the first cat population, represents the initial position of the i-th cat in the second cat population;

[0049] Start the iterative operation, set the current iteration number of the first cat population as n1′, and the current iteration number of the second cat population as n′2; calculate the fitness of each cat in the first cat population and the second cat population during each round of iteration, and obtain the fitness set of the first cat population and the fitness set of the second cat population and respectively represent the fitness value of the i-th cat in the first cat population and the fitness value of the i-th cat in the second cat population;

[0050] According to the fitness set of the first cat population and the fitness set of the second cat population obtain the optimal cat individual in the first cat population and the optimal cat individual in the second cat population in the current round of iteration;

[0051] S27. Randomly set the sub - cat population n1 that executes the search mode in the first cat population souxun and the sub - cat population n1 that executes the tracking mode genzong as well as the sub - cat population n2 that executes the search mode in the second cat population souxun and the sub - cat population n2 that executes the tracking mode genzong ;

[0052] Then, update the positions of each cat in the first cat population according to n1 souxun and n1 genzong ; update the positions of each cat in the second cat population according to n2 souxun and n2 genzong ;

[0053] S28. Update the first initial classification center set and the second initial classification center set l′ = {l1′, l2′,..., l i ′,..., l h ′} using the nearest neighbor rule, and update the fitness set of the first cat population and the fitness set of the second cat population ;

[0054] After S29, the optimal category set of customer priority screening factor data and the optimal category set of customer elimination screening factor data are obtained;

[0055] The cat swarm algorithm is used to classify the data of the two types of customers collected and continuously optimize the categories. Finally, the optimal classification is obtained. In the specific process of classifying the customer data, the algorithm models under the two behavior patterns in the cat swarm algorithm are fully utilized, making the convergence speed of the cat swarm algorithm faster when classifying customer data, and it is not easy to fall into the local optimum situation, improving the classification efficiency, and ultimately improving the efficiency of screening customers.

[0056] Preferably, the fitness calculation formula for each cat in the first cat swarm population in S26 is as follows:

[0057]

[0058] In the formula, o 1i represents the fitness of the i-th cat in the first cat swarm population, and o1′ i represents the sum of the dispersions of the i-th initial customer priority screening factor classification data;

[0059] The fitness calculation formula for each cat in the second cat swarm population is as follows,

[0060]

[0061] In the formula, o 2i represents the fitness of the i-th cat in the second cat swarm population, and o′ 2i represents the sum of the dispersions of the i-th customer elimination screening factor classification data;

[0062] By calculating the fitness, the optimization goal is determined, making the classification direction more accurate.

[0063] Preferably, the specific process of S29 is as follows:

[0064] When the iteration process of the first cat swarm population ends, and the optimal category set of customer priority screening factor data is output

[0065] When the iteration process of the second cat swarm population ends, and the optimal category set of customer elimination screening factor data is output

[0066] and respectively represent the optimal category data set of the i-th customer priority screening factor data and the optimal category data set of the i-th customer elimination screening factor data;

[0067] When the condition for iterative stop is reached, stop the optimization, and at this time, output the optimal classification.

[0068] Preferably, the step S3 includes the following steps:

[0069] S31. Set the priority screening classification threshold and the elimination screening classification threshold

[0070] S32. Perform category ranking on the optimal category set of customer priority screening factor data and the optimal category set of customer elimination screening factor data to obtain the sorted optimal category set of customer priority screening factor data and the sorted optimal category set of customer elimination screening factor data;

[0071] S33. Retain the customers in the top categories of the sorted optimal category set of customer priority screening factor data; eliminate the customers in the top categories of the sorted optimal category set of customer elimination screening factor data; obtain the final customer set p = {p1, p2,..., p i ,..., p p′}, where p i represents the i-th customer finally screened out;

[0072] By ranking the categories of customers, it is convenient for enterprise decision-makers to perform customer screening and elimination operations, and finally obtain the highest-quality customers.

[0073] Preferably, the step S4 includes the following steps:

[0074] S41. Collect the corresponding final customer advertisement type matrix according to the final customer set p = {p1, p2,..., p i ,..., p p′} as follows, as follows,

[0075]

[0076] where represents the j-th advertisement type corresponding to the i-th customer in the final customer set; represents the total number of advertisement types of the i-th customer in the final customer set;

[0077] S42. Set the advertisement set q to be screened = {q1, q2,..., q i ,..., q q′}, where q idenote the i-th advertisement to be screened, and q' denote the total number of advertisements to be screened; collect the types of each advertisement to be screened in the advertisement set to be screened q = {q1, q2,..., q i ,..., q q′}, and obtain the advertisement type set to be screened denote the type of the i-th advertisement to be screened;

[0078] S43. Compare the advertisement type set to be screened with the final customer advertisement type matrix to obtain the advertisement recommendation set denote the i-th screened advertisement, r denote the total number of screened advertisements, and finally recommend the advertisement recommendation set to the enterprise decision maker;

[0079] The selection of advertisement types is based on the advertisement types already screened and issued by the customers themselves, so that the advertisement method has stronger pertinence and higher efficiency.

[0080] A customer screening and advertisement recommendation system based on data analysis, including a customer priority screening factor data collection module, a customer elimination screening factor data collection module, a customer priority screening factor data classification module, a customer elimination screening factor data classification module, a customer priority screening factor data category optimization module, a customer elimination screening factor data category optimization module, a sorting module, a screening and elimination module, and an advertisement screening module;

[0081] The customer priority screening factor data collection module is used to collect the customer priority screening factor data to be screened according to the customer set to be screened and the customer priority screening factor set;

[0082] The customer elimination screening factor data collection module is used to collect the customer elimination screening factor data to be screened according to the customer set to be screened and the customer elimination screening factor set;

[0083] The customer priority screening factor data classification module is used to classify the customer priority screening factor data to be screened;

[0084] The customer elimination screening factor data classification module is used to classify the customer elimination screening factor data to be screened;

[0085] The customer priority screening factor data category optimization module is used to classify and optimize the first customer weighted data category matrix to be screened;

[0086] The customer elimination screening factor data category optimization module is used to classify and optimize the second customer weighted data category matrix to be screened;

[0087] The sorting module is used to sort the optimal category sets of customer priority screening factor data and customer elimination screening factor data;

[0088] The screening and elimination module is used to screen and eliminate customers in the sorted optimal category sets of customer priority screening factor data and customer elimination screening factor data according to the priority screening classification threshold and the elimination screening classification threshold respectively;

[0089] The advertisement screening module is used to compare the to-be-screened advertisement type set with the final customer advertisement type matrix to obtain an advertisement recommendation set.

[0090] The present invention has the following beneficial effects:

[0091] 1. In the present invention, by setting a customer priority screening factor data acquisition module, a customer elimination screening factor data acquisition module, a customer priority screening factor data classification module, a customer elimination screening factor data classification module, a customer priority screening factor data category optimization module, a customer elimination screening factor data category optimization module, a sorting module, a screening and elimination module, and an advertisement screening module; the customer priority screening factor set includes, for example, consumption habits, browsing history, social media activities, the needs, preferences, behavior patterns, purchase probability, and consumption ability of the customer's target market, etc., and the customer elimination screening factor set includes, for example, the customer's bad credit record and whether there is a criminal record, etc.; by collecting data of each customer according to the above influencing factors, the information of the customer can be comprehensively collected from multiple aspects, making the subsequent analysis more comprehensive and accurate; classifying the collected data one by one, and then overall classifying the already classified category data, so as to determine the category of the customer under the comprehensive influence of the above factors, so that the fit degree between the customer and the enterprise and the goodness or badness of the customer himself can be separated according to the category, and then sorting the customers according to the category of the customer, and the customers can be conveniently screened according to the sorting result; since the advertisement is often sent to the customer, in the solution of the present invention, the selection of the advertisement type is based on the advertisement type already sent by the already screened customer himself, so that the method of the advertisement has stronger pertinence and higher efficiency; by adopting the above solution, only the customer information data needs to be collected and the data is input into the system for classification, without too much human participation, improving the objectivity of the screening and reducing the screening cost of the enterprise.

[0092] 2. In the present invention, by setting the customer priority screening factor set and the customer elimination screening factor set, it provides a collection direction for subsequent customer information collection, making the collection of customer information more targeted and comprehensive; then, by collecting the first customer data matrix to be screened and the second customer data matrix to be screened, it provides data support for subsequent customer classification; by classifying each data in the first customer data matrix to be screened and the second customer data matrix to be screened, it makes the subsequent overall customer classification more convenient and efficient; among them, since the importance of each influencing factor is different, through weighted operation, the data accuracy is improved, thus also improving the accuracy of subsequent customer classification.

[0093] 3. In the present invention, by using the cat swarm algorithm to classify the data of two types of customers collected and continuously optimize the categories, the optimal classification is finally obtained. In the specific process of classifying the customer data, the algorithm models under the two behavior patterns in the cat swarm algorithm are fully utilized, making the convergence speed of the cat swarm algorithm faster when classifying customer data, and not easily falling into the local optimum situation, improving the classification efficiency, and ultimately improving the efficiency of customer screening.

[0094] Of course, it is not necessary for any product implementing the present invention to achieve all the above-mentioned advantages simultaneously. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] In order to more clearly illustrate the technical solutions of the embodiments of the invention, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings without creative efforts.

[0096] Figure 1 It is a schematic flowchart of a customer screening and advertisement recommendation system based on data analysis according to the present invention for screening customers and advertisements. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0097] The technical solutions in the embodiments of the invention will be clearly and completely described below with reference to the drawings in the embodiments of the invention. Obviously, the described embodiments are only a part of the embodiments of the invention, rather than all the embodiments. Based on the embodiments of the invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the invention.

[0098] In the description of the present invention, it should be understood that the terms "open hole", "upper", "lower", "top", "middle", "inner", etc. indicating orientation or positional relationship are only for the convenience of describing the invention and simplifying the description, rather than indicating or implying that the components or elements referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation of the invention.

[0099] The present invention is a method for customer screening and advertisement recommendation based on data analysis, comprising the following steps:

[0100] S1. Set a customer priority screening factor set and a customer elimination screening factor set; set a set of customers to be screened; collect data on customer priority screening factors for the customers to be screened according to the set of customers to be screened and the customer priority screening factor set; then collect data on customer elimination screening factors for the customers to be screened according to the set of customers to be screened and the customer elimination screening factor set; classify the data on customer priority screening factors and data on customer elimination screening factors for the customers to be screened, and finally obtain a first weighted data category matrix of customers to be screened and a second weighted data category matrix of customers to be screened;

[0101] The said S1 comprises the following steps:

[0102] S11. Set a customer priority screening factor set \(a = \{a_1, a_2, \cdots, a_j, \cdots, a_{a'}\}\) and a customer elimination screening factor set j , \cdots, a a′ \}, where \(a_j\) represents the \(j\)th customer priority screening factor and \(a'\) represents the total number of set customer priority screening factors; a j represents the \(j\)th customer elimination screening factor, represents the total number of set customer priority screening factors; set a set of customers to be screened \(b = \{b_1, b_2, \cdots, b_i, \cdots, b_{b'}\}\), where \(b_i\) represents the \(i\)th customer to be screened and \(b'\) represents the total number of set customers to be screened; i , \cdots, b b′ \}, b i represents the \(i\)th customer to be screened and \(b'\) represents the total number of set customers to be screened;

[0103] S12. Collect data on customer priority screening factors for the set of customers to be screened \(b = \{b_1, b_2, \cdots, b_i, \cdots, b_{b'}\}\) and the customer priority screening factor set \(a = \{a_1, a_2, \cdots, a_j, \cdots, a_{a'}\}\) to obtain a first data matrix of customers to be screened i , \cdots, b b′ j , \cdots, a, \cdots, a a′ as follows, as follows,

[0104]

[0105] wherein, represents the data of the i-th customer to be screened in the j-th customer priority screening factor;

[0106] Then, according to the set of customers to be screened b = {b1, b2,..., b i ,..., b b′} and the set of customer elimination screening factors collect the data of the customer elimination screening factors of the customers to be screened to obtain the second matrix of data of customers to be screened as follows,

[0107]

[0108] wherein, represents the data of the i-th customer to be screened in the j-th customer elimination screening factor;

[0109] S13. Set the set of customer priority screening levels c = {c1, c2,..., c i ,..., c c′}, where c i represents the i-th customer priority screening level, and c' represents the total number of set customer priority screening levels; according to the set of customer priority screening factors a = {a1, a2,..., a j ,..., a a′} and the set of customer priority screening levels c = {c1, c2,..., c i ,..., c c′} set the customer priority screening level data interval matrix as follows

[0110]

[0111] wherein, and respectively represent the upper and lower limits of the data interval of the i-th customer priority screening level on the j-th customer priority screening factor;

[0112] Set the set of customer elimination screening levels represents the i-th customer elimination screening level, and d represents the total number of set customer elimination screening levels; according to the set of customer elimination screening factors and the set of customer elimination screening levels set the customer elimination screening level data interval matrix d'; as follows,

[0113]

[0114] wherein, and respectively represent the upper and lower limits of the data interval of the i-th customer elimination and screening level on the j-th customer elimination and screening factor;

[0115] S14. According to the customer priority screening level data interval matrix classify the data of the first customer to be screened matrix to obtain the first customer to be screened data category matrix as follows,

[0116]

[0117] wherein, represents the data category of the i-th customer to be screened in the j-th customer priority screening factor;

[0118] Then, according to the customer elimination and screening level data interval matrix d′, classify the data of the second customer to be screened matrix to obtain the second customer to be screened data category matrix as follows,

[0119]

[0120] wherein, represents the data category of the i-th customer to be screened in the j-th customer elimination and screening factor;

[0121] S15. Set the customer priority screening factor weight set f = {f1, f2,... f j ,..., f a′} and the customer elimination and screening factor weight set f j represents the weight value of the j-th customer priority screening factor; f i ′ represents the weight value of the i-th customer elimination and screening factor;

[0122] According to the customer priority screening factor weight set f = {f1, f2,... f j ,..., f a′} and the first customer to be screened data category matrix obtain the first customer to be screened weighted data category matrix Then, according to the customer elimination and screening factor weight set and the second customer to be screened data category matrix obtain the second customer to be screened weighted data category matrix The first customer to be screened weighted data category matrix and the second customer to be screened weighted data category matrix are respectively as follows,

[0123]

[0124] Among them, represents the weighted data category of the i-th customer to be screened in the j-th customer priority screening factor; represents the weighted data category of the i-th customer to be screened in the j-th customer elimination screening factor;

[0125] The first weighted data category matrix of customers to be screened described in S15 and the second weighted data category matrix of customers to be screened The calculation formulas for the data in are as follows:

[0126]

[0127] S2. Construct a first cat population and a second cat population to classify and optimize the first weighted data category matrix of customers to be screened and the second weighted data category matrix of customers to be screened respectively, and obtain an optimal category set of customer priority screening factor data and an optimal category set of customer elimination screening factor data;

[0128] S2 includes the following steps:

[0129] S21. Construct a first cat population and a second cat population, set the scale of the first cat population as e, the scale of the second cat population as g′, and the first cat population and the second cat population are respectively expressed as e′ = {e1′, e′2,..., e i ′,..., e e ′} and e i ′ represents the i-th cat in the first cat population, represents the i-th cat in the second cat population; set the first maximum number of iterations as The second maximum number of iterations is Set the number of position space dimensions of the cats in the first cat population as The number of position space dimensions of the cats in the second cat population is h; set the grouping rate of the first cat population as m, and the grouping rate of the second cat population as m′;

[0130] S22. According to the number of position space dimensions of the cats in the first cat population, set the total category set of customer priority screening factor data g = {g1, g2,..., g i ,..., g e～}, g i represents the i-th total category of customer priority screening factor data; then, according to the number of position space dimensions h of the cats in the second cat population, set the total category set of customer elimination screening factor data h′ = {h1′, h2′,..., hi ′,...,h h ′}, h i ′ represents the total category of the data of the i-th customer elimination screening factor;

[0131] S23. Randomly classify each data in the first customer weighted data category matrix into the set of total categories of customer priority screening factor data g = {g1, g2,..., g i ,..., g e～}, and obtain the initial customer priority screening factor classification data set Then randomly classify each data in the second customer weighted data category matrix into the set of total categories of customer elimination screening factor data h′ = {h1′, h2′,..., h i ′,..., h h ′}, and obtain the initial customer elimination screening factor classification data set represents the i-th initial customer priority screening factor classification data, represents the i-th initial customer elimination screening factor classification data;

[0132] S24. Calculate the classification centers of each classification data in the initial customer priority screening factor classification data set and the initial customer elimination screening factor classification data set respectively, and obtain the first initial classification center set and the second initial classification center set l′ = {l1′, l2′,..., l i ′,..., l h ′}; l i represents 's classification center, l i ′ represents 's classification center;

[0133] S25. Take the first initial classification center set and the second initial classification center set l′ = {l1′, l2′,..., l i ′,..., l h ′} as the position of the first cat e1′ in the first cat population and the position of the first cat in the second cat population respectively;

[0134] S26. Repeat S23, S24, and S25, and finally obtain the initial position set

[0135] of the first cat population and the initial position set represents the initial position of the i-th cat in the first cat population, represents the initial position of the i-th cat in the second cat population;

[0136] Start the iterative operation, set the current iteration number of the first cat population as n1′, and the current iteration number of the second cat population as n′2; calculate the fitness of each cat in the first cat population and the second cat population during each round of iteration, and respectively obtain the fitness set of the first cat population and the fitness set of the second cat population and respectively represent the fitness value of the i-th cat in the first cat population and the fitness value of the i-th cat in the second cat population;

[0137] According to the fitness set of the first cat population and the fitness set of the second cat population, obtain the optimal cat individual in the first cat population and the optimal cat individual in the second cat population in the current round of iteration;

[0138] The fitness calculation formula for each cat in the first cat population in S26 is as follows:

[0139]

[0140] In the formula, o 1i represents the fitness of the i-th cat in the first cat population, and o1′ i represents the sum of the dispersions of the i-th initial customer priority screening factor classification data;

[0141] The fitness calculation formula for each cat in the second cat population is as follows,

[0142]

[0143] In the formula, o 2i represents the fitness of the i-th cat in the second cat population, and o′ 2i represents the sum of the dispersions of the i-th customer elimination screening factor classification data;

[0144] S27. Randomly set the sub-cat group n1 souxun executing the search mode and the sub-cat group n1 genzong executing the tracking mode in the first cat population, as well as the sub-cat group n2 souxun executing the search mode and the sub-cat group n2 genzong executing the tracking mode in the second cat population according to the grouping rate m of the first cat population and the grouping rate m′ of the second cat population;

[0145] Then, according to n1 souxun and n1 genzongUpdate the positions of each cat in the first cat population; according to n2 souxun and n2 genzong Update the positions of each cat in the second cat population;

[0146] S28. Use the nearest neighbor rule to update the first initial classification center set and the second initial classification center set l′ = {l1′, l2′,..., l i ′,..., l h ′}, and update the fitness set of the first cat population and the fitness set of the second cat population ;

[0147] S29. After the iteration ends, obtain the optimal category set of customer priority screening factor data and the optimal category set of customer elimination screening factor data;

[0148] The specific process of S29 is as follows:

[0149] When the iteration process of the first cat population ends, output the optimal category set of customer priority screening factor data

[0150] When the iteration process of the second cat population ends, output the optimal category set of customer elimination screening factor data

[0151] and respectively represent the optimal category data set of the i-th customer priority screening factor and the optimal category data set of the i-th customer elimination screening factor;

[0152] S3. Set the priority screening classification threshold and the elimination screening classification threshold; sort the categories of the optimal category set of customer priority screening factor data and the optimal category set of customer elimination screening factor data, and screen and eliminate the customers in the sorted optimal category set of customer priority screening factor data and the optimal category set of customer elimination screening factor data according to the priority screening classification threshold and the elimination screening classification threshold to obtain the final customer set;

[0153] S3 includes the following steps:

[0154] S31. Set the priority screening classification threshold and the elimination screening classification threshold

[0155] S32. For the optimal category set of customer priority screening factor data And the optimal category set of customer elimination screening factor data Perform category sorting to obtain the sorted optimal category set of customer priority screening factor data and the sorted optimal category set of customer elimination screening factor data;

[0156] S33. Retain the customers in the first categories of the sorted optimal category set of customer priority screening factor data; Eliminate the customers in the first categories of the sorted optimal category set of customer elimination screening factor data; Obtain the final customer set p = {p1, p2,..., p i ,..., p p′}, where p i represents the i-th customer finally screened out;

[0157] S4. Collect the corresponding final customer advertisement type matrix according to the final customer set; Set the advertisement set to be screened and collect the types of each advertisement to be screened in the advertisement set to be screened to obtain the advertisement type set to be screened. Compare the advertisement type set to be screened with the final customer advertisement type matrix to obtain the advertisement recommendation set; Finally, recommend the advertisement recommendation set to the enterprise decision-maker;

[0158] The S4 includes the following steps:

[0159] S41. Collect the corresponding final customer advertisement type matrix according to the final customer set p = {p1, p2,..., p i ,..., p p′} as follows, where,

[0160]

[0161] Among them, represents the j-th advertisement type corresponding to the i-th customer in the final customer set; represents the total number of advertisement types of the i-th customer in the final customer set;

[0162] S42. Set the advertisement set to be screened q = {q1, q2,..., q i ,..., q q′}, where q i represents the i-th advertisement to be screened, and q' represents the total number of advertisements to be screened; Collect the types of each advertisement to be screened in the advertisement set to be screened q = {q1, q2,..., q i ,..., q q′} to obtain the advertisement type set to be screened represents the type of the i-th advertisement to be screened;

[0163] S43. Compare the set of advertising types to be screened with the matrix of advertising types of end customers to obtain an advertising recommendation set Let the i-th screened advertisement be represented, and r represents the total number of screened advertisements. Finally, recommend the advertising recommendation set to the enterprise decision maker.

[0164] A customer screening and advertising recommendation system based on data analysis, including a customer priority screening factor data acquisition module, a customer elimination screening factor data acquisition module, a customer priority screening factor data classification module, a customer elimination screening factor data classification module, a customer priority screening factor data category optimization module, a customer elimination screening factor data category optimization module, a sorting module, a screening and elimination module, and an advertising screening module;

[0165] The customer priority screening factor data acquisition module is used to collect the data of the customer priority screening factors to be screened according to the set of customers to be screened and the set of customer priority screening factors;

[0166] The customer elimination screening factor data acquisition module is used to collect the data of the customer elimination screening factors to be screened according to the set of customers to be screened and the set of customer elimination screening factors;

[0167] The customer priority screening factor data classification module is used to classify the data of the customer priority screening factors to be screened;

[0168] The customer elimination screening factor data classification module is used to classify the data of the customer elimination screening factors to be screened;

[0169] The customer priority screening factor data category optimization module is used to optimize the classification of the first matrix of weighted data categories of customers to be screened;

[0170] The customer elimination screening factor data category optimization module is used to optimize the classification of the second matrix of weighted data categories of customers to be screened;

[0171] The sorting module is used to sort the optimal category sets of the customer priority screening factor data and the optimal category sets of the customer elimination screening factor data;

[0172] The screening and elimination module is used to screen and eliminate the customers in the optimal category sets of the customer priority screening factor data and the optimal category sets of the customer elimination screening factor data after sorting according to the priority screening classification threshold and the elimination screening classification threshold;

[0173] The advertising screening module is used to compare the set of advertising types to be screened with the matrix of advertising types of end customers to obtain an advertising recommendation set.

[0174] In the description of this specification, the descriptions referring to the terms "one embodiment", "example", "specific example", etc. mean that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0175] The preferred embodiments of the invention disclosed above are only used to help illustrate the invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific embodiments described. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification in order to better explain the principle and practical application of the invention, so that those skilled in the art can well understand and utilize the invention.

Claims

1. A customer screening and advertisement recommendation method based on data analysis, characterized in that, It includes the following steps: S1. Set the customer priority screening factor set and the customer elimination screening factor set; Set the customer set to be screened; collect the customer priority screening factor data of the customers to be screened according to the customer set to be screened and the customer priority screening factor set; Then collect the customer elimination screening factor data of the customers to be screened according to the customer set to be screened and the customer elimination screening factor set; Classify the customer priority screening factor data and the customer elimination screening factor data of the customers to be screened, and finally obtain the first weighted data category matrix of the customers to be screened and the second weighted data category matrix of the customers to be screened; S2. Construct a first cat population and a second cat population to classify and optimize the first weighted data category matrix of the customers to be screened and the second weighted data category matrix of the customers to be screened respectively, and obtain the optimal category set of the customer priority screening factor data and the optimal category set of the customer elimination screening factor data; S3. Set the priority screening classification threshold and the elimination screening classification threshold; Rank the optimal category set of the customer priority screening factor data and the optimal category set of the customer elimination screening factor data, and screen and eliminate the customers in the sorted optimal category set of the customer priority screening factor data and the optimal category set of the customer elimination screening factor data according to the priority screening classification threshold and the elimination screening classification threshold to obtain the final customer set; S4. Collect the corresponding final customer advertisement type matrix according to the final customer set; Set the advertisement set to be screened and collect the types of each advertisement to be screened in the advertisement set to be screened to obtain the advertisement type set to be screened, and compare the advertisement type set to be screened with the final customer advertisement type matrix to obtain the advertisement recommendation set; The S1 includes the following steps: S11. Set the customer priority screening factor set and the customer elimination screening factor set ; denote the th customer priority screening factor, denote the total number of set customer priority screening factors; denote the th customer elimination screening factor, denote the total number of set customer priority screening factors; Set the customer set to be screened , denote the th customer to be screened, denote the total number of set customers to be screened; S12. According to the to-be-screened customer set and the customer priority screening factor set collect the to-be-screened customer priority screening factor data to obtain the first to-be-screened customer data matrix as follows ； Among them, represents the data of the th customer to be screened in the th customer priority screening factor; According to the customer set to be screened and the customer elimination screening factor set Collect the data of the customer elimination screening factors to be screened, and obtain the second customer data matrix to be screened ; as follows ； Among them, represents the data of the th customer to be screened in the th customer elimination screening factor; S13. Set the customer priority screening level set , represents the th customer priority screening level, represents the total number of set customer priority screening levels; according to the customer priority screening factor set and the customer priority screening level set set the customer priority screening level data interval matrix ; as follows ； Among them, and respectively represent the upper and lower limits of the data range of the th customer priority screening level on the th customer priority screening factor; Set the customer elimination screening level set , represents the th customer elimination screening level, represents the total number of the set customer elimination screening levels; according to the customer elimination screening factor set and the customer elimination screening level set set the customer elimination screening level data interval matrix ; as follows, ； Among them, and respectively represent the upper and lower limits of the data range of the -th customer elimination screening level on the -th customer elimination screening factor; S14. According to the customer priority screening level data interval matrix classify the first customer data matrix to be screened to obtain the first customer data category matrix to be screened ; as follows ； Among them, represents the data category of the th customer to be screened in the th customer priority screening factor; According to the customer elimination and screening level data interval matrix classify the second customer data matrix to be screened to obtain the second customer data category matrix to be screened ; as follows ； Among them, represents the data category of the th customer to be screened in the th customer elimination screening factor; S15. Set the weight set of customer priority screening factors and the weight set of customer elimination screening factors ; represents the weight value of the th customer priority screening factor; represents the weight value of the th customer elimination screening factor; According to the weight set of customer priority screening factors and the first customer data category matrix to be screened obtain the first weighted customer data category matrix to be screened , and then according to the weight set of customer elimination screening factors and the second customer data category matrix to be screened obtain the second weighted customer data category matrix to be screened , the first weighted customer data category matrix to be screened and the second weighted customer data category matrix to be screened are as follows respectively ； ； Among them, represents the weighted data category of the th customer to be screened in the th customer priority screening factor; represents the weighted data category of the th customer to be screened in the th customer elimination screening factor.

2. The method for customer screening and advertisement recommendation based on data analysis according to claim 1, characterized in that The first customer weighted data category matrix described in S15 and the second customer weighted data category matrix The calculation formulas for the data in them are as follows: ； 。 3. A method for customer screening and advertisement recommendation based on data analysis according to claim 1, characterized in that The S2 includes the following steps: S21. Construct the first cat population and the second cat population, and set the size of the first cat population as , and the size of the second cat population as . The first cat population and the second cat population are respectively represented as and ; represents the -th cat in the first cat population, represents the -th cat in the second cat population; set the first maximum iteration number as , the second maximum iteration number as ; set the number of position space dimensions of the cats in the first cat population as , and the number of position space dimensions of the cats in the second cat population as ; set the grouping rate of the first cat population as , and the grouping rate of the second cat population as ; S22. According to the number of position space dimensions of the cats in the first cat population Set the total category set of customer priority screening factor data , Denote the th total category of customer priority screening factor data; then according to the number of position space dimensions of the cats in the second cat population Set the total category set of customer elimination screening factor data , Denote the th total category of customer elimination screening factor data; S23. Randomly classify each data in the first customer weighted data category matrix into the total customer priority screening factor data category set to obtain an initial customer priority screening factor classification data set ; then randomly classify each data in the second customer weighted data category matrix into the total customer elimination screening factor data category set to obtain an initial customer elimination screening factor classification data set ; represents the th initial customer priority screening factor classification data, represents the th initial customer elimination screening factor classification data; S24. Calculate the initial customer priority screening factor classification data set and the initial customer elimination screening factor classification data set to obtain the first initial classification center set and the second initial classification center set respectively; Denote as the classification center of Denote as the classification center of; S25. Take the first initial classification center set and the second initial classification center set as the position of the first cat in the first cat population and the position of the first cat in the second cat population respectively; S26. Repeat S23, S24, and S25 to finally obtain the initial position set of the first cat population and the initial position set of the second cat population ; denotes the initial position of the -th cat in the first cat population, denotes the initial position of the -th cat in the second cat population; Start the iterative operation, and set the current iteration number of the first cat population as , and the current iteration number of the second cat population as ; During each iteration process, calculate the fitness of each cat in the first cat population and the second cat population, and obtain the first cat population fitness set and the second cat population fitness set ; and respectively represent the fitness value of the -th cat in the first cat population and the fitness value of the -th cat in the second cat population; According to the first cat group population fitness set and the second cat group population fitness set obtain the optimal cat individual in the first cat group population and the optimal cat individual in the second cat group population in the current round of iteration; S27. According to the first cat population grouping rate and the second cat population grouping rate randomly set the sub - cat groups in the first cat population that execute the search mode and the sub - cat groups that execute the tracking mode as well as the sub - cat groups in the second cat population that execute the search mode and the sub - cat groups that execute the tracking mode ; Then, according to and update the positions of the cats in the first cat population; according to and update the positions of the cats in the second cat population; S28. Update the first initial classification center set and the second initial classification center set using the nearest neighbor rule, and update the first cat swarm population fitness set and the second cat swarm population fitness set ; S29. After the iteration ends, obtain the optimal category set of the customer priority screening factor data and the optimal category set of the customer elimination screening factor data.

4. A method for customer screening and advertisement recommendation based on data analysis according to claim 3, characterized in that, The fitness calculation formula of each cat in the first cat population in S26 is as follows: ； In the formula, represents the fitness of the th cat in the first cat population, represents the sum of the dispersions of the th initial customer priority screening factor classification data; The fitness calculation formula of each cat in the second cat population is as follows, ； In the formula, represents the fitness of the th cat in the second cat population, represents the sum of the dispersions of the classification data of the th customer elimination and screening factor.

5. A method for customer screening and advertisement recommendation based on data analysis according to claim 3, characterized in that, The specific process of the S29 is as follows: When the iterative process of the first cat group population ends, output the optimal category set of customer priority screening factor data ; When the iterative process of the second cat population ends, output the optimal category set of customer elimination and screening factor data ; and respectively represent the optimal category data set of the th customer priority screening factor and the optimal category data set of the th customer elimination screening factor.

6. A method for customer screening and advertisement recommendation based on data analysis according to claim 1, characterized in that The S3 includes the following steps: S31. Set the priority screening classification threshold and the elimination screening classification threshold ; S32. Optimize the category sets of the customer priority screening factor data and the optimized category set of the customer elimination screening factor data Perform category sorting to obtain the sorted optimized category set of the customer priority screening factor data and the sorted optimized category set of the customer elimination screening factor data; S33. Retain the customers in the top number of categories in the optimal category set of the sorted customer priority screening factor data; eliminate the customers in the top number of categories in the optimal category set of the sorted customer elimination screening factor data; obtain the final customer set , indicating the th customer finally screened out.

7. A method for customer screening and advertisement recommendation based on data analysis according to claim 1, characterized in that The S4 includes the following steps: S41. According to the final customer set collect the corresponding final customer advertisement type matrix ; as follows ； Among them, represents the th advertising type corresponding to the th customer in the end-customer set; represents the total number of advertising types of the th customer in the end-customer set; S42. Set the advertising set to be screened , indicating the th advertising to be screened, and representing the total number of advertisements to be screened; Collect the types of each advertisement to be screened in the advertising set to be screened to obtain an advertising type set to be screened , indicating the type of the th advertisement to be screened; S43. Compare the set of advertising types to be screened with the final customer advertising type matrix to obtain an advertising recommendation set , indicating the th screened advertisement, and representing the total number of screened advertisements.

8. A system for implementing the method for customer screening and advertisement recommendation based on data analysis according to any one of claims 1-7, characterized in that: It includes a customer priority screening factor data collection module, a customer elimination screening factor data collection module, a customer priority screening factor data classification module, a customer elimination screening factor data classification module, a customer priority screening factor data category optimization module, a customer elimination screening factor data category optimization module, a sorting module, a screening and elimination module, and an advertisement screening module; The customer priority screening factor data collection module is used to collect the customer priority screening factor data of the customers to be screened according to the customer set to be screened and the customer priority screening factor set; The customer elimination screening factor data collection module is used to collect the customer elimination screening factor data of the customers to be screened according to the customer set to be screened and the customer elimination screening factor set; The customer priority screening factor data classification module is used to classify the customer priority screening factor data of the customers to be screened; The customer elimination screening factor data classification module is used to classify the customer elimination screening factor data of the customers to be screened; The customer priority screening factor data category optimization module is used to classify and optimize the first customer weighted data category matrix to be screened; The customer elimination screening factor data category optimization module is used to classify and optimize the second customer weighted data category matrix to be screened; The sorting module is used to sort the optimal category sets of the customer priority screening factor data and the customer elimination screening factor data; The screening and elimination module is used to screen and eliminate the customers in the sorted optimal category sets of the customer priority screening factor data and the customer elimination screening factor data according to the priority screening classification threshold and the elimination screening classification threshold respectively; The advertisement screening module is used to compare the advertisement type set to be screened with the final customer advertisement type matrix to obtain an advertisement recommendation set.

Images