Big Data-Based Customer Satisfaction Analysis Method and System

Abstract

The present invention provides a customer satisfaction analysis method and system based on big data, belonging to the field of data analysis; it solves the problem of low efficiency in customer satisfaction analysis; specifically as follows: Step S1: Obtain order data; Step S2: Calculate the sales coefficient of each product according to the order data; according to the questionnaire information and the sales coefficient, analyze the customer satisfaction with each product, and determine the popular products and unpopular products; Step S3: Obtain the product type of the popular products as the quasi-type; obtain the order data of the quasi-type products sold by different enterprises, use SARIMA regression to analyze the order number differentiation of the popular products in different seasons, and formulate a sales plan for the popular products; Step S4: Summarize the unpopular products and the sales plan and give feedback; The present invention analyzes the order data and questionnaire data of different types of products, analyzes the customer satisfaction with each product from a relatively objective perspective, and provides a reference for enterprise production.

Description

Technical Field

[0001] The customer satisfaction analysis method and system based on big data of the present invention relate to the field of data analysis. Background Art

[0002] The existing methods or systems for customer satisfaction analysis have the following deficiencies:

[0003] Data errors and anomalies: There may be errors and outliers in customer satisfaction data, such as filling errors, malicious evaluations, etc.; these errors and outliers will interfere with and affect the data analysis results, leading to incorrect decisions and conclusions, and a data processing method for multi-dimensional information extraction is required.

[0004] High technical threshold: The existing customer satisfaction analysis systems involve complex technologies and algorithms, such as data mining, machine learning, natural language processing, etc.; these technologies require professional technical personnel for development and maintenance, posing relatively high requirements for the technical capabilities of enterprises.

[0005] High application cost: Building a customer satisfaction analysis system requires a large investment in hardware devices, such as servers, storage devices, network devices, etc.; the procurement, installation, and maintenance costs of these hardware devices are relatively high; at the same time, during the process of customer satisfaction analysis, enterprises need a large number of data statistics talents, and the labor cost is relatively high. Summary of the Invention

[0006] Aiming at the deficiencies of the existing technology, the purpose of the present invention is to provide a customer satisfaction analysis method and system based on big data, aiming to solve the problem of low efficiency of customer satisfaction analysis.

[0007] In order to achieve the above purpose, the present invention is realized through the following technical solutions: The customer satisfaction analysis method based on big data includes:

[0008] Step S1: Obtain the number of product types, and obtain the order quantity of each product type as order data;

[0009] Step S2: Calculate the sales coefficient of each product according to the order data; collect questionnaire information of different customers on each product; analyze the customer satisfaction of each product according to the questionnaire information and the sales coefficient, and determine the popular products and unpopular products;

[0010] Step S3: Obtain the product type of the popular product as the quasi-type; obtain the order data of different enterprises selling the quasi-type products, use SARIMA regression to analyze the order quantity differentiation of the popular product in different seasons, and formulate a sales plan for the popular product;

[0011] Step S4: Summarize the unpopular products and the sales plan and give feedback.

[0012] Furthermore, the specific steps of step S2 are as follows:

[0013] Step S21: Denote the total number of products as pr, and the total number of days in the past three months as td;

[0014] Denote the order numbers of the first type of product from the 1st to the td-th day as og （1,1） ~og （td,1） ;

[0015] Similarly, denote the order numbers of the pr-th type of product from the 1st to the td-th day as og （1,pr） ~og （td,pr） ;

[0016] Step S22: Calculate the average value aog of og （1,1） ~og （td,1） as the average order number of the first type of product; （1） Similarly, the average value aog of og

[0017] ~og （1,pr） ~og （td,pr） is used as the average order number of the pr-th type of product; （pr） Define an empty matrix of (td×pr), and fill it with og

[0018] （1,1） ~og~og （td,pr） to obtain matrix BO:

[0019] ;

[0020] Calculate the sales coefficients of the first to the pr-th types of products according to matrix BO, and obtain bu （1） ~bu （pr） ;

[0021] Step S23: Define scoring indicators: product quality, product price, after-sales service, arrival time;

[0022] Count the product quality scores of the first to the pr-th types of products given by the first to the us-th customers, denoted as: ra （1,1） ~ra （us,pr） ; product price scores, denoted as: rb （1,1） ~rb （us,pr） ; after-sales service scores, denoted as: rc （1,1） ~rc （us,pr） ; arrival time scores, denoted as: rd （1,1） ~rd （us,pr） where us represents the number of customers participating in the questionnaire survey;

[0023] Step S24: Extract the product quality scores ra of the first type of product given by the first to the us-th customers（1,1） ~ra （us,1） ,product price score rb （1,1） ~rb （us,1） ,after-sales service score rc （1,1） ~rc （us,1） ,arrival time score rd （1,1） ~rd （us,1） ;

[0024] Calculate the recognition coefficient ac of the first product （1） .

[0025] Furthermore, the step S2 further includes:

[0026] Step S25: Repeatedly calculate ac （1） using the same steps, calculate the recognition coefficients of the second to the pr-th products, and obtain ac （2） ~ac （pr） ; Calculate the average value aac of ac （1） ~ac （pr） ;

[0027] Obtain the sales coefficients bu of the first to the pr-th products （1） ~bu （pr） ;

[0028] Define the relational expression A-8-1:

[0029] ; bu （m） and bu （n） , respectively represent the sales coefficients of the m-th and the n-th products;

[0030] Mark the products that satisfy the relational expression A-8-1 as high-sales products;

[0031] Step S26: Obtain the recognition coefficients gac of the high-sales products （1） ~gac （rs） ; rs represents the number of high-sales products;

[0032] Define the relational expression A-8-2: gac （w） ≥aac; gac （w） represents the recognition coefficient of the w-th high-sales product;

[0033] Count the high-sales products that satisfy the relational expression A-8-2 as best-selling products;

[0034] Obtain the recognition coefficients eac of the best-selling products （1） ~eac （re） ; re represents the number of best-selling products;

[0035] Step S27: Obtain the sales coefficient bl of the remaining products （1） ~bl （pi） , recognition coefficient oac （1） ~oac （pi） ; pi represents the quantity of the remaining products;

[0036] Extract the maximum value blm and the minimum value bll in bl （1） ~bl （pi） ; eac （1） ~eac （re） The maximum value eam and the minimum value eal in;

[0037] Define the calculation formula A-8-3:

[0038] ;

[0039] cr （f） 、bl （f） and oac （f） , respectively represent the satisfaction coefficient, sales coefficient and recognition coefficient of the f-th remaining product;

[0040] Calculate the satisfaction coefficient cr of the remaining products from the 1st to the pi-th （1） ~cr （pi） ;

[0041] Extract the products with a satisfaction coefficient less than 0 as unpopular products, and the products with a satisfaction coefficient greater than 1 as popular products.

[0042] Furthermore, the specific steps of the said step S22 are as follows:

[0043] Step S221: Denote the average order number of the i-th product as aog （i） , and the order number of the i-th product on the j-th day as og （j，i） ;

[0044] Define the calculation formula A-1:

[0045] ; ag （j，i） represents the first-order standardized value of og （j，i） ;

[0046] Calculate the first-order standardized values of all elements in the matrix BO to obtain the matrix BO （1） :

[0047] ;

[0048] Step S222: Calculate the average value Cag of all elements in the first column of the matrix BO （1） in （1）; The average value Cag of all elements in the second column （2） ; Similarly, the average value Cag of all elements in the pr-th column （pr） ;

[0049] Define calculation formula A-2:

[0050] ; SC （i） represents the standard deviation of all elements in the i-th column of matrix BO （1） ;

[0051] Calculate the standard deviation SC of all elements in the first column （1） , the standard deviation SC of all elements in the second column （2） ; Similarly, the standard deviation SC of all elements in the pr-th column （pr） ;

[0052] Step S223: Define calculation formula A-3:

[0053] ; bg （j，i） represents the second-order standardized value of ag （j，i） ;

[0054] Calculate the second-order standardized values of all elements in matrix BO （1） to obtain matrix BO （2） :

[0055] ;

[0056] Define calculation formula A-4-1: ;

[0057] Define calculation formula A-4-2: ;

[0058] Rbg （j） represents the average value of all elements in the j-th row of matrix BO （2） , Cbg （i） represents the average value of all elements in the i-th column of matrix BO （2） ;

[0059] Calculate the average values Rbg （2） of all elements from the 1st to the td-th row in matrix BO （1） ~Rbg （td） , and the average values Cbg （1） ~Cbg （pr） .

[0060] Furthermore, the subsequent steps of step S223 are as follows:

[0061] Step S224: Define calculation formula A-5:

[0062] ; among which, cg （j，i） represents the third-order normalized value of bg （j，i） ;

[0063] Calculate the third-order normalized values of all elements in matrix BO （2） to obtain matrix BO （3） :

[0064] ;

[0065] Split matrix BO （3） by columns to obtain matrices CBO （1） , CBO （2） ~ matrix CBO （pr） ;

[0066] Matrix CBO （1） is: ; Matrix CBO （2） is: ;

[0067] And so on, matrix CBO （pr） is: ;

[0068] Step S225: Define the parameter coefficients λ （1） , λ （2） ~ λ （pr） and the offset coefficients μ （1） , μ （2） ~ μ （pr） and μ （1） , μ （2） ~ μ （pr） for matrices CBO

[0069] Use the Intel MKL library function to calculate λ （1） ~ λ （pr） and μ （1） ~ μ （pr） so that matrices CBO （1） ~ matrix CBO （pr） satisfy the following mathematical relationship:

[0070] ;

[0071] Among which, II （td） represents the all-ones matrix of (td×1);

[0072] Step S226: Calculate the sales coefficient bu （1） of the first type of product, bu （1） = λ （1） / μ（1） ;

[0073] The sales coefficient bu of the second product （2） , bu （2） = λ （2） / μ （2） ;

[0074] And so on, the sales coefficient bu of the pr-th product （pr） , bu （pr） = λ （pr） / μ （pr） .

[0075] Furthermore, the specific steps of the said step S24 are as follows:

[0076] Step S241: Calculate the total score ara of the first customer （1） ; Similarly, the total score ara of the us-th customer （us） ; Calculate the sum Ara of ara （1） ~ ara （us） ;

[0077] Step S242: According to ra （1,1） ~ ra （us,1） , calculate the weighted coefficients λa （1） ~ λa （us） and the offset coefficient va of the product quality scores of the first to the us-th customers;

[0078] Step S243: Repeatedly calculate the same steps of λa （1） ~ λa （us） and va, and calculate the weighted coefficients λb （1） ~ λb （us） and the offset coefficient vb of the product price scores of the first to the us-th customers;

[0079] Calculate the weighted coefficients λc （1） ~ λc （us） and the offset coefficient vc of the after-sales service scores;

[0080] Calculate the weighted coefficients λd （1） ~ λd （us） and the offset coefficient vd of the arrival time scores;

[0081] Step S244: Corresponding to the empty matrix of (us × 4), and fill in ra （1,1） ~ ra （us,1） , rb （1,1） ~ rb （us,1） , rc （1,1） ~ rc （us,1） and rd （1,1） ~ rd （us,1）, obtain matrix CO:

[0082] ;

[0083] Divide matrix CO by column to obtain matrix CA, matrix CB, matrix CC, and matrix CD;

[0084] Matrix CA: ; Matrix CB: ; Matrix CC: ;

[0085] Matrix CD: .

[0086] Furthermore, the subsequent steps of step S244 are as follows:

[0087] Step S245: Calculate the path coefficient β of the product quality score corresponding to the first product （ra） : ; where T represents the transpose of a matrix, × represents matrix multiplication, and -1 represents the inverse of a matrix;

[0088] Path coefficient β of the product price score （rb） : ;

[0089] Path coefficient β of the after-sales service score （rc） : ;

[0090] Path coefficient β of the arrival time score （rd） : ;

[0091] Step S246: Calculate the recognition coefficient ac of the first product （1） ;

[0092] Define calculation formula A-7:

[0093] ;

[0094] where λa （t） , λb （t） , λc （t） and λd （t） respectively represent the weighted coefficients of the t-th customer for the product quality score, product price score, after-sales service score, and arrival time score.

[0095] Furthermore, the specific steps of step S242 are as follows:

[0096] Step S2421: Assume that the values of λa （1） ~λa （us） after k updates are ar （1）（k） ~ar （us） （k） ;

[0097] Calculate ra （1,1） The value of raa after k updates （1,1） （k） , raa （1,1） （k） = ra （1,1） × ar （1） （k） ;

[0098] Similarly, ra （us,1） The value of raa after k updates （us,1） （k） , raa （us,1） （k） = ra （us,1） × ar （us） （k） ;

[0099] Step S2422: Calculate ra （1,1） ~ ra （us,1） The k - weighted sum of, denoted as SQ （k） :

[0100] ; where, ra （t,1） Represents the product quality score of the t - th customer for the first product, ar （t） （k） Represents the value of the weighted coefficient of the product quality score of the t - th customer after k updates; the value range of t is: 1 ~ us;

[0101] Step S2423: Calculate the mean value denoted as rajj and the variance denoted as ss of ra （1,1） ~ ra （us,1） after k updates （k） , variance denoted as ss （ra） （k） ;

[0102] Calculate the mean value denoted as raj of ra （1,1） ~ ra （us,1） ; According to raj and rajj, calculate the k - covariance denoted as cov of ra （1,1） ~ ra （us,1） with respect to raa （1,1） ~ raa （us,1） denoted as cov （ra） （k） ;

[0103] Define calculation formula A - 6 - 1:

[0104] ; where, raa （t,1）（k） Denote the k-th updated value of ra （t,1） as Aλ （k） which represents the combined weight after k updates.

[0105] Furthermore, the subsequent steps of step S2423 are as follows:

[0106] Step S2424: Set the initial values of ar （1） to ar （us） as 0.5;

[0107] Set the value of k to 1 and update Aλ （k） until the relational expression A-6-2 is satisfied;

[0108] The relational expression A-6-2 is: ; where Aλ （k+1） represents the combined weight after (k + 1) updates; ε represents the allowable error value;

[0109] Denote the combined weight that satisfies the relational expression A-6-2 as Bλ;

[0110] Step S2425: The weighting coefficient λa of the product quality score of the first customer （1） is:

[0111] ;

[0112] The weighting coefficient λa of the product quality score of the second customer （2） is:

[0113] ;

[0114] And so on, the weighting coefficient λa of the product quality score of the us-th customer （us） is:

[0115] ;

[0116] Step S2426: Calculate the offset coefficient va and define the calculation formula A-6-3:

[0117] ; λa （t） represents the weighting coefficient of the product quality score of the t-th customer.

[0118] The customer satisfaction analysis system based on big data includes:

[0119] Order acquisition module: used to acquire the types of products and the number of orders for each product as order data;

[0120] Data analysis module: satisfaction analysis sub-module and sales analysis sub-module;

[0121] Satisfaction analysis sub-module: used to calculate the sales coefficient of each product according to the order data; collect questionnaire information of different customers for each product; analyze the customer satisfaction of each product based on the questionnaire information and sales coefficient, and determine the popular products and unpopular products;

[0122] Sales analysis sub-module: used to obtain the product type of the popular product as the quasi-type; obtain the order data of the quasi-type product sold by different enterprises, use SARIMA regression to analyze the differentiation of the order quantity of the popular product in different seasons, and formulate the sales plan of the popular product;

[0123] User interaction module: used to summarize the unpopular products and sales plans and give feedback.

[0124] Compared with the prior art, the beneficial effects of the present invention are:

[0125] Precisely identify the root cause of the problem: By analyzing the relevant data of the perceived orders and questionnaires, the enterprise can dig out the key factors affecting customer satisfaction, such as product quality problems, unsmooth service processes, unreasonable prices, etc.; this precise problem positioning helps the enterprise formulate targeted improvement measures, thereby improving the scientificity and accuracy of decision-making.

[0126] Predict customer behavior trends: The big data analysis of the present invention not only focuses on the current data, but also can predict future customer behavior trends through the analysis of historical data, enabling the enterprise to make advance arrangements to meet the future needs of customers and enhance customer satisfaction and loyalty.

[0127] Continuously improve the service process: By continuously analyzing the customer satisfaction data, the enterprise can discover the bottlenecks and inefficient links in the service process and optimize them accordingly. This continuously improved service process can enhance the overall customer experience and the competitiveness of the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0128] By reading the following detailed description of the non-limiting embodiments with reference to the accompanying drawings, other features, objects and advantages of the present invention will become more apparent:

[0129] Figure 1 It is a schematic diagram of the method of the present invention;

[0130] Figure 2 It is a schematic diagram of the system of the present invention;

[0131] Figure 3 It is a schematic diagram of the questionnaire survey of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0132] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0133] Embodiment 1:

[0134] Please refer to Figure 1 and Figure 3 , the customer satisfaction analysis method based on big data includes:

[0135] Step S1: Obtain the number of types of products (provided by the enterprise), and obtain the number of orders for each product (in the recent three months) as order data;

[0136] Step S2: Calculate the sales coefficient of each product; (Distribute online or offline questionnaires for each product to customers) Collect questionnaire information of different customers for each product; Analyze the customer satisfaction of each product based on the questionnaire information and the sales coefficient, and determine popular products and unpopular products;

[0137] The specific steps of Step S2 are as follows:

[0138] Step S21: Denote the total number of products as pr; Take the number of orders for each product (in the recent three months) as order data; Denote the total number of days in the recent three months as td;

[0139] Denote the number of orders of the first product on the 1st, 2nd, up to the td-th day as og （1,1） 、og （2,1） ~og （td,1） ;

[0140] Denote the number of orders of the second product on the 1st, 2nd, up to the td-th day as og （1,2） 、og （2,2） ~og （td,2） ;

[0141] And so on, denote the number of orders of the pr-th product on the 1st, 2nd, up to the td-th day as og （1,pr） 、og （2,pr） ~og （td,pr） ;

[0142] It should be noted that the "order" in the present invention refers to an "effective order", that is, an order that has completed a complete transaction; If an order is withdrawn or interrupted by a consumer during the transaction process, then this order is not included in the "effective order".

[0143] Step S22: Calculate the average value aog （1,1） ~og （td,1） as the average number of orders for the first product; og （1） ~og （1,2） ~og （td,2）The average value aog （1） As the average number of orders for the second product; and so on, og （1,pr） ~og （td,pr） The average value aog （pr） As the average number of orders for the pr-th product;

[0144] Define an empty matrix of (td × pr) and fill it with og （1,1） ~og （td,pr） to obtain matrix BO:

[0145] ;

[0146] Calculate the sales coefficients of the first to pr-th products based on matrix BO to obtain bu （1） ~bu （pr） ;

[0147] Step S221: Denote the average number of orders for the i-th product as aog （i） , and denote the number of orders for the i-th product on the j-th day as og （j，i） ; The value range of i is: 1~pr, and the value range of j is: 1~td;

[0148] Define calculation formula A-1:

[0149] ; Among them, ag （j，i） represents the first-order standardized value of og （j，i） ;

[0150] Calculate the first-order standardized values of all elements in matrix BO (according to calculation formula A-1) to obtain matrix BO （1） :

[0151] ;

[0152] Step S222: Calculate the average value of all elements in the first column of matrix BO （1） (that is, ag （1,1） ~ag （td,1） ), denoted as Cag （1） ; The average value of all elements in the second column (that is, ag （1,2） ~ag （td,2） ), denoted as Cag （2） ; And so on, the average value of all elements in the pr-th column (that is, ag （1,pr） ~ag （td,pr） ), denoted as Cag （pr） ;

[0153] Define calculation formula A-2:

[0154] ; Among them, SC（i） Denote the matrix BO （1） the standard deviation of all elements in the i-th column of

[0155] (According to calculation formula A-2), calculate the standard deviation SC of all elements in the first column (i.e., ag （1,1） ~ag （td,1） ); the standard deviation SC of all elements in the second column (i.e., ag （1） ~ag （1,2） ~ag （td,2） ); and so on, the standard deviation SC of all elements in the pr-th column (i.e., ag （2） ~ag （1,pr） ~ag （td,pr） ); （pr） ;

[0156] Step S223: Define calculation formula A-3:

[0157] ; where bg （j，i） represents the second-order standardized value of ag （j，i） ;

[0158] (According to calculation formula A-3), calculate the second-order standardized values of all elements in the matrix BO （1） to obtain the matrix BO （2） :

[0159] ;

[0160] Define calculation formula A-4-1: ;

[0161] Define calculation formula A-4-2: ;

[0162] where Rbg （j） represents the average value of all elements in the j-th row of the matrix BO （2） , and Cbg （i） represents the average value of all elements in the i-th column of the matrix BO （2） ;

[0163] (According to calculation formula A-4-1 and calculation formula A-4-1), calculate the average values Rbg （2） ~Rbg （1） ~Rbg （td） of all elements from the 1st to the td-th row of the matrix BO （1） ~Cbg （pr） of all elements from the 1st to the pr-th column;

[0164] Step S224: Define calculation formula A-5:

[0165] ; where, cg （j，i） represents the cubic normalization value of bg （j，i） ;

[0166] (According to calculation formula A-5) Calculate the cubic normalization values of all elements in matrix BO （2） to obtain matrix BO （3） :

[0167] ;

[0168] Split matrix BO （3） by column to obtain matrix CBO （1） , CBO （2） ~matrix CBO （pr） ;

[0169] Matrix CBO （1） is: ; Matrix CBO （2） is: ;

[0170] And so on, matrix CBO （pr） is: ;

[0171] Step S225: Define the parameter coefficients λ （1） , CBO （2） ~matrix CBO （pr） and the offset coefficients μ （1） , λ （2） ~λ （pr） and μ （1） , μ （2） ~μ （pr） ;

[0172] Use the Intel MKL library function to calculate λ （1） ~λ （pr） and μ （1） ~μ （pr） to make matrix CBO （1） ~matrix CBO （pr） satisfy the following mathematical relationship:

[0173] ;

[0174] where, II （td） represents the all-one matrix of (td×1);

[0175] Step S226: Calculate the sales coefficient bu （1） of the first product, bu （1） =λ （1） / μ （1） ;

[0176] Sales coefficient bu of the second product （2） , bu （2） = λ （2） / μ （2） ;

[0177] And so on, the sales coefficient bu of the pr-th product （pr） , bu （pr） = λ （pr） / μ （pr） ;

[0178] Step S23: Define the scoring metrics: product quality, product price, after-sales service, arrival time; (the upper limit of the score is 10 points)

[0179] Send a questionnaire (online or offline for the 1st to pr-th products) to us customers; us represents the number of customers participating in the questionnaire survey

[0180] Statistical product quality scores of the 1st to us customers for the 1st to pr-th products, denoted as: ra （1,1） ~ra （us,pr） ; Product price scores, denoted as: rb （1,1） ~rb （us,pr） ; After-sales service scores, denoted as: rc （1,1） ~rc （us,pr） ; Arrival time scores, denoted as: rd （1,1） ~rd （us,pr） ;

[0181] Step S24: Extract the product quality scores ra （1,1） , ra （2,1） ~ra （us,1） , product price scores rb （1,1） , rb （2,1） ~rb （us,1） , after-sales service scores rc （1,1） , rc （2,1） ~rc （us,1） , arrival time scores rd （1,1） , rd （2,1） ~rd （us,1） ;

[0182] Calculate the recognition coefficient ac of the 1st product （1） ;

[0183] Step S241: Calculate the total score of the 1st customer (i.e., the sum of ra （1,1） , rb （1,1） , rc （1,1） and rd （1,1） ), denoted as ara （1）;

[0184] The total score of the second customer (i.e., the sum of ra （2,1） , rb （2,1） , rc （2,1） and rd （2,1） ) is denoted as ara （2） ;

[0185] And so on, the total score of the us-th customer (i.e., the sum of ra （us,1） , rb （us,1） , rc （us,1） and rd （us,1） ) is denoted as ara （us） ;

[0186] Calculate the sum of ara （1） to ara （us） , and denote it as Ara;

[0187] Step S242: According to ra （1,1） to ra （us,1） , calculate the weighted coefficients λa （1） to λa （us） and the offset coefficient va of the product quality scores of the 1st to us-th customers for the (first product);

[0188] Step S2421: Assume that the values of λa （1） to λa （us） after the k-th update are ar （1） （k） to ar （us） （k） ; (k is a natural number)

[0189] Calculate the value raa （1,1） （1,1） （k） of ra （1,1） after the k-th update, raa （1,1） （k） = ra （1,1） ×ar （1） （k） ;

[0190] The value raa （2,1） （2,1） （k） of ra （2,1） （k） after the k-th update, raa （2,1） ×ar （2） （k） ;

[0191] And so on, the value raa （us,1） of ra （us,1） （k） after the k-th update, raa（us,1） （k） =ra （us,1） × （us） （k） ;

[0192] Step S2422: (based on ar （1） （k） ~ar （us） （k） ) Calculate ra （1,1） ~ra （us,1） The k-times weighted sum of （k） :

[0193] ; Among them, ra （t,1） represents the calculation of the product quality score of the t-th customer for the first product, ar （t） （k） It represents the value of the weighted coefficient of the product quality score of the t-th customer (for the first product) after k updates; the value range of t is: 1~us;

[0194] Step S2423: Calculate ra （1,1） ~ra （us,1） (ie raa （1,1） （k） ~raa （us,1） （k） ) The corresponding k-times updated mean is denoted as rajj （k） , the variance is recorded as ss （ra） （k） ;

[0195] Calculate ra （1,1） ~ra （us,1） The mean of is denoted as raj; based on raj and rajj, calculate ra （1,1） ~ra （us,1） About raa （1,1） ~raa （us,1） The k-th covariance of （ra） （k） ;

[0196] Define calculation formula A-6-1:

[0197] ; Among them, raa （t,1） （k） Indicates ra （t,1） k-times updated value, Aλ （k） Represents the combined weight after k updates;

[0198] Step S2424: ar （1） ~ar （us）The initial value is set to 0.5; (Customers can adjust the size of the initial value according to actual needs)

[0199] Set the value of k to 1, and update Aλ according to the rules corresponding to steps S2421 to S2423 （k） until the relational expression A-6-2 is satisfied;

[0200] The relational expression A-6-2 is: ; where Aλ （k+1） represents the combined weight after (k + 1) updates; ε represents the allowable error value; (The value of ε is 0.01, and customers or relevant technical personnel can adjust the value of ε according to actual needs)

[0201] Record the combined weight that satisfies the relational expression A-6-2 as Bλ;

[0202] Step S2425: The weighting coefficient λa of the product quality score of the first customer (for the first product) （1） is:

[0203] ;

[0204] The weighting coefficient λa of the product quality score of the second customer (for the first product) （2） is:

[0205] ;

[0206] And so on, the weighting coefficient λa of the product quality score of the us-th customer (for the first product) （us） is:

[0207] ;

[0208] Step S2426: Define the calculation formula A-6-3:

[0209] ; where λa （t） represents the weighting coefficient of the product quality score of the t-th customer (for the first product);

[0210] Calculate the offset coefficient va according to the calculation formula A-6-3;

[0211] Step S243: Repeat the same steps of calculating λa （1） ~λa （us） and va to calculate the weighting coefficients λb （1） ~λb （us） of the product price scores of the first to us-th customers (for the first product) and the offset coefficient vb;

[0212] Weighting coefficient λc for the after-sales service score of (the first product) （1） ~λc （us） and offset coefficient vc;

[0213] Weighting coefficient λd for the arrival time score of (the first product) （1） ~λd （us） and offset coefficient vd;

[0214] Step S244: Corresponding to an empty matrix of (us×4), and sequentially fill in ra （1,1） ~ra （us,1） , rb （1,1） ~rb （us,1） , rc （1,1） ~rc （us,1） and rd （1,1） ~rd （us,1） , to obtain matrix CO:

[0215] ;

[0216] Divide matrix CO by columns to obtain matrix CA, matrix CB, matrix CC, and matrix CD;

[0217] Matrix CA: ; Matrix CB: ; Matrix CC: ;

[0218] Matrix CD: ;

[0219] Step S245: Calculate the path coefficient β of the product quality score corresponding to the first product (the overall score of the first product) （ra） : ; where T represents the transpose of the matrix, × represents matrix multiplication, and -1 represents the inverse of the matrix;

[0220] Path coefficient β of the product price score (the overall score of the first product) （rb） : ;

[0221] Path coefficient β of the after-sales service score (the overall score of the first product) （rc） : ;

[0222] Path coefficient β of the arrival time score (the overall score of the first product) （rd） : ;

[0223] Step S246: Calculate the recognition coefficient ac of the first product (according to calculation formula A-7) （1） ;

[0224] Define calculation formula A-7:

[0225] ;

[0226] where λa （t） , λb （t） , λc （t） and λd （t） respectively represent the weighted coefficients of the t-th customer's ratings on product quality, product price, after-sales service, and arrival time of the first product;

[0227] Step S25: Repeat the same steps of calculating ac （1） to calculate the recognition coefficients of the 2nd to the pr-th products, obtaining ac （2） to ac （pr） ; calculate the average value in ac （1） to ac （pr） , denoted as aac;

[0228] Obtain the sales coefficients bu （1） to bu （pr） of the 1st to the pr-th products;

[0229] Define relation A-8-1:

[0230] ; where bu （m） and bu （n） respectively represent the sales coefficients of the m-th and n-th products; m and n ∈ {1~pr} and m ≠ n;

[0231] (According to bu （1） to bu （pr） ) Mark the products that satisfy relation A-8-1 as high-sales products;

[0232] Step S26: Obtain the recognition coefficients of the high-sales products, denoted as gac （1） to gac （rs） ; where rs represents the number of high-sales products;

[0233] Define relation A-8-2: gac （w） ≥ aac; where gac （w） represents the recognition coefficient of the w-th high-sales product; the value range of w is: 1~rs;

[0234] (According to gac （1） to gac （rs） ) Count the high-sales products that satisfy relation A-8-2 as best-selling products;

[0235] Obtain the recognition coefficients of the best-selling products, denoted as eac（1） ~eac （re） ; where re represents the quantity of best-selling products;

[0236] Step S27: (Among the 1st to pr-th products, eliminate the best-selling products to obtain the remaining products)

[0237] (Among bu （1） ~bu （pr） and in ac （1） ~ac （pr） ), eliminate the sales coefficient and recognition coefficient corresponding to the popular high-sales products to obtain bl （1） ~bl （pi） and oac （1） ~oac （pi） ; where pi represents the quantity of the remaining products (after eliminating the best-selling products), pi = pr - re;

[0238] Extract the maximum value in bl （1） ~bl （pi） and denote it as blm, and the minimum value as bll;

[0239] Extract the maximum value in eac （1） ~eac （re） and denote it as eam, and the minimum value as eal;

[0240] Define calculation formula A-8-3:

[0241] ;

[0242] where cr （f） , bl （f） and oac （f） , respectively represent the satisfaction coefficient, sales coefficient and recognition coefficient of the f-th remaining product; the value range of f is: 1~pi;

[0243] (According to calculation formula A-8-3) Calculate the satisfaction coefficient of the 1st to pi-th remaining products to obtain cr （1） ~cr （pi） ;

[0244] Among the 1st to pi-th remaining products, extract the products with a satisfaction coefficient less than 0 as unpopular products, and the products with a satisfaction coefficient close to or greater than 1 as popular products, (a satisfaction coefficient close to 1 means the satisfaction coefficient ≥ 0.97)

[0245] Take the popular products and the best-selling products as the hot products;

[0246] Step S3: Obtain the product type of popular products as the quasi-type; obtain the order data of different enterprises (in the past three years) selling products of the quasi-type, use SARIMA regression to analyze the differentiation of the order numbers of popular products in different seasons, and formulate the sales plan for popular products;

[0247] It should be noted that the "quasi-type" in the present invention represents the commodity type;

[0248] For example, enterprise A is an enterprise using the present invention. If the popular product of enterprise A is ** brand smartphones, then the quasi-type is "smartphones"; the "order data" obtained in step S3 is: the order data (or sales data) of smartphones of other enterprises (enterprises other than enterprise A) in the past three years on the market;

[0249] Step S4: Summarize the sales plans of unpopular products and (popular products) and give feedback.

[0250] It should be noted that the "product quality, product price, after-sales service, arrival time" in the present invention are only for sample demonstrations, and customers or relevant technical personnel can define or adjust more product scoring indicators according to actual needs.

[0251] Actual example two:

[0252] Please refer to Figure 2 , the customer satisfaction analysis system based on big data includes:

[0253] Order acquisition module: used to obtain the types of products (provided by the enterprise), and obtain the order numbers of each product (in the past three months) as the order data;

[0254] Data analysis module: satisfaction analysis sub-module and sales analysis sub-module;

[0255] Satisfaction analysis sub-module: used to calculate the sales coefficient of each product according to the order data; (distribute online or offline questionnaires for each product to customers) collect the questionnaire information of different customers for each product; analyze the satisfaction of customers with each product according to the questionnaire information and the sales coefficient, and determine popular products and unpopular products;

[0256] Sales analysis sub-module: used to obtain the product type of popular products as the quasi-type; obtain the order data of different enterprises (in the past three years) selling products of the quasi-type, use SARIMA regression to analyze the differentiation of the order numbers of popular products in different seasons, and formulate the sales plan for popular products;

[0257] User interaction module: used to summarize the sales plans of unpopular products and (popular products) and give feedback.

[0258] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation. For example, there are weight coefficients and proportionality coefficients, and the sizes of their settings are for obtaining a specific numerical value by quantifying each parameter, which is convenient for subsequent comparison. Regarding the sizes of the weight coefficients and proportionality coefficients, as long as they do not affect the proportional relationship between the parameters and the quantified numerical values, it is acceptable.

[0259] Finally, it should be noted that the above-described embodiments are only specific implementation manners of the present invention, used to illustrate the technical solutions of the present invention, rather than limiting it. The protection scope of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: any person skilled in the art within the technical scope disclosed by the present invention can still modify the technical solutions recorded in the foregoing embodiments, or can easily think of changes, or perform equivalent replacements on some of the technical features; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A method for analyzing customer satisfaction based on big data, characterized in that, The method includes: Step S1: Obtain the number of product types, and obtain the order quantity of each product as order data; Step S2: Calculate the sales coefficient of each product according to the order data; collect questionnaire information of different customers for each product; according to the questionnaire information and the sales coefficient, analyze the satisfaction of customers with each product, and determine popular products and unpopular products; Extract the product quality score ra of the first product （1,1） ~ra （us,1） , the product price score rb （1,1） ~rb （us,1） , the after-sales service score rc （1,1） ~rc （us,1） , the arrival time score rd （1,1） ~rd （us,1） , calculate the recognition coefficient ac of the first product （1）； Step S241: Calculate the total score ara of the 1st to the us-th customer （1） ~ara （us） ; Calculate ara （1） ~ara （us） and Ara; Step S242: Calculate the weighted coefficients λa （1） （1） ~λa （us） （us） and the offset coefficient va for the product quality scores given by the 1st to the us-th customers according to the product quality scores. Let λa after k updates （1） ~λa （us） have the value ar （1） （k） ~ar （us） （k） ; Calculate ra （1,1） The value raa after k updates （1,1） （k） = ra （1,1） × ar （1） （k） ; Similarly, ra （us,1） The value raa after k updates （us,1） （k） ; Calculate ra （1,1） ~ra （us,1） The k - weighted sum SQ of ra （k） ; Calculate ra （1,1） ~ra （us,1） The mean value after the k-th update corresponding to it is denoted as rajj （k） , and the variance is denoted as ss （ra） （k） ; Calculate ra （1,1） ~ra （us,1） The mean value of is denoted as raj; according to raj and rajj, calculate ra （1,1） ~ra （us,1） Regarding raa （1,1） ~raa （us,1） The k-th covariance of is denoted as cov （ra） （k） ; ; wherein, raa （t,1） （k） represents the k-th updated value of ra （t,1） , and Aλ （k） represents the combined weight after k updates; Step S243: Calculate the weighted coefficient and offset coefficient vb of the product price score of the 1st to the us-th customers; calculate the weighted coefficient and offset coefficient vc of the after-sales service score; calculate the weighted coefficient and offset coefficient vd of the arrival time score; Step S244: For the empty matrix of (us × 4), and fill in ra （1,1） ~ra （us,1） , rb （1,1） ~rb （us,1） , rc （1,1） ~rc （us,1） and rd （1,1） ~rd （us,1） , to obtain matrix CO: Divide the matrix CO by columns to obtain matrix CA, matrix CB, matrix CC, and matrix CD; Step S245: Calculate the path coefficient β of the product quality score corresponding to the first type of product （ra） : ; where T represents the transpose of a matrix, × represents matrix multiplication, and -1 represents the inverse of a matrix; Path coefficient of product price scoring ; Path coefficient of after-sales service score ; Path coefficient of arrival time score ; Step S246: Calculate the recognition coefficient ac of the first type of product （1） ; ； λa （t） , λb （t） ,λc （t） and λd （t） , respectively represent the weighted coefficients of the t-th customer’s ratings on product quality, product price, after-sales service, and delivery time; Step S3: Obtain the product type of the popular product as the quasi-type; obtain the order data of the quasi-type products sold by different enterprises, and use SARIMA regression to analyze the order quantity differentiation of the popular products in different seasons, and formulate a sales plan for the popular products; Step S4: Summarize the unpopular products and the sales plan and give feedback.

2. The method for analyzing customer satisfaction based on big data according to claim 1, wherein The specific steps of the said Step S2 are as follows: Step S21: Denote the total number of products as pr, and denote the total number of days in the past three months as td; Record the number of orders for the first product from the 1st day to the td-th day as og （1,1） ~og （td,1） ; Similarly, denote the order numbers of the pr-th product from the 1st day to the td-th day as og （1,pr） ~og （td,pr） ; Step S22: Calculate og （1,1） ~og （td,1） The average value aog of （1） is used as the average order quantity of the first type of product; Similarly, og （1,pr） ~og （td,pr） The average value aog （pr） is used as the average number of orders for the pr-th product; Define an empty matrix of (td × pr) and fill it with og （1,1） ~og （td,pr） , to obtain matrix BO: ； Calculate the sales coefficients of the 1st to prth products according to matrix BO to obtain bu （1） ~bu （pr） ; Step S23: Define the scoring indicators: product quality, product price, after-sales service, arrival time; Statistically evaluate the product quality scores of the 1st to the us-th customers for the 1st to the pr-th products, denoted as: ra （1,1） ~ra （us,pr） ; product price scores, denoted as: rb （1,1） ~rb （us,pr） ; after-sales service scores, denoted as: rc （1,1） ~rc （us,pr） ; arrival time scores, denoted as: rd （1,1） ~rd （us,pr） ; us represents the number of customers participating in the questionnaire survey; Step S24: Extract the product quality scores ra of the 1st to the us-th customers for the 1st product （1,1） ~ra （us,1） , the product price scores rb （1,1） ~rb （us,1） , the after-sales service scores rc （1,1） ~rc （us,1） , the arrival time scores rd （1,1） ~rd （us,1） ; Calculate the recognition coefficient ac of the first product （1） .

3. The method for analyzing customer satisfaction based on big data according to claim 2, wherein The said Step S2 also includes: Step S25: Repeatedly calculate ac （1） for the same steps to calculate the recognition coefficients of the 2nd to the pr-th products, obtaining ac （2） ~ac （pr） ; calculate the average value aac of ac （1） ~ac （pr） ; Obtain the sales coefficients bu of the 1st to the prth products （1） ~bu （pr） ; Define the relation A-8-1: ; bu （m） and bu （n） , representing the sales coefficients of the m-th and n-th products respectively; Mark the products that satisfy the relation A-8-1 as high-sales products; Step S26: Obtain the recognition coefficient gac of high-sales products （1） ~gac （rs） ; rs represents the quantity of high-sales products; Define the relational expression A-8-2: gac （w） ≥aac; gac （w） represents the recognition coefficient of the w-th high-sales product Count the high-sales products that satisfy the relation A-8-2 as best-selling products; Obtain the recognition coefficient eac of best-selling products （1） ~eac （re） ; re represents the quantity of best-selling products; Step S27: Obtain the sales coefficient bl of the remaining products （1） ~bl （pi） , and the recognition coefficient oac （1） ~oac （pi） ; pi represents the quantity of the remaining products Extract bl （1） ~bl （pi） The maximum value blm and the minimum value bll therein; eac （1） ~eac （re） The maximum value eam and the minimum value eal therein; Define the calculation formula A-8-3: ； cr （f） ,bl （f） and oac （f） , respectively represent the satisfaction coefficient, sales coefficient and recognition coefficient of the f-th remaining product; Calculate the satisfaction coefficient cr of the first to the pi-th remaining products （1） ~cr （pi） ; Extract the products with a satisfaction coefficient less than 0 as unpopular products, and the products with a satisfaction coefficient greater than 1 as popular products.

4. The method for analyzing customer satisfaction based on big data according to claim 2, wherein The specific steps of the said Step S22 are as follows: Step S221: Denote the average order quantity of the i-th product as aog （i） , and denote the order quantity of the i-th product on the j-th day as og （j，i） ; Define the calculation formula A-1: ; ag （j，i） represents og （j，i） 's first standardized value; Calculate the first normalization value of all elements in matrix BO to obtain matrix BO （1） : ； Step S222: Calculate matrix BO （1） The average value Cag of all elements in the first column of （1） ; Similarly, the average value Cag of all elements in the pr-th column （pr） ; Define the calculation formula A-2: ; SC （i） represents the standard deviation of all elements in the i-th column of matrix BO （1） in Calculate the standard deviation SC of all elements in the first column （1） to the standard deviation SC of all elements in the pr-th column （pr） ; Step S223: Define the calculation formula A-3: ; bg （j，i） represents ag （j，i） 's secondary normalization value; Calculate the quadratic normalization values of all elements in matrix BO （1） to obtain matrix BO （2） : ； Define calculation formula A-4-1: ; Define calculation formula A-4-2: ; Rbg （j） represents the average value of all elements in the j-th row of matrix BO （2） , and Cbg （i） represents the average value of all elements in the i-th column of matrix BO （2） ; Calculate matrix BO （2） The average value Rbg of all elements in the 1st to the td-th rows （1） ~Rbg （td） and the average value Cbg of all elements in the 1st to the pr-th columns （1） ~Cbg （pr） .

5. The method for analyzing customer satisfaction based on big data according to claim 4, wherein The subsequent steps of the said Step S223 are as follows: Step S224: Define the calculation formula A-5: ; wherein, cg （j，i） represents the triple normalization value of bg （j，i） ; Calculate the cubic standardized values of all elements in matrix BO （2） to obtain matrix BO （3） : ； Split the matrix BO （3） by columns to obtain the matrix CBO （1） ~ matrix CBO （pr） ; Matrix CBO （1） is: ; and so on, matrix CBO （pr） is: ; Step S225: Define matrix CBO （1） ~matrix CBO （pr） with parameter coefficient λ （1） ~λ （pr） and offset coefficient μ （1） ~μ （pr） ; Calculate λ using the Intel MKL library function （1） ~λ （pr） and μ （1） ~μ （pr） such that matrix CBO （1） ~matrix CBO （pr） satisfies the following mathematical relationship: ； Among them, II （td） represents a matrix of all ones of (td×1); Step S226: Calculate the sales coefficient bu of the first type of product （1） , bu （1） = λ （1） / μ （1） ; Sales coefficient bu of the second product （2） , bu （2） = λ （2） / μ （2） ; And so on, the sales coefficient bu of the pr-th product （pr） , bu （pr） = λ （pr） / μ （pr） .

6. The method for analyzing customer satisfaction based on big data according to claim 1, characterized in that The said Step S242 also includes: Set the initial value of ar （1） ~ ar （us） to 0.5; Set the value of k to 1 and update Aλ （k） until the relational expression A-6-2 is satisfied; The relational expression A-6-2 is as follows: ; where, Aλ （k+1） represents the combined weight after the (k + 1)-th update; ε represents the allowable error value; Denote the combined weight that satisfies the relation A-6-2 as Bλ; The weighted coefficient λa of the product quality score of the first customer （1） is as follows: ； By analogy, the weighted coefficient λa of the product quality score of the us-th customer （us） is as follows: ； Calculate the offset coefficient va, and define the calculation formula A-6-3: ; λa （t） represents the weighted coefficient of the product quality score of the t-th customer.

7. A customer satisfaction analysis system based on big data, applicable to the big data-based customer satisfaction analysis method described in any one of claims 1-6, characterized in that, The system includes: An order acquisition module: used to obtain the number of product types, and obtain the order quantity of each product as order data; A data analysis module: a satisfaction analysis sub-module and a sales analysis sub-module; The satisfaction analysis sub-module: used to calculate the sales coefficient of each product according to the order data; collect questionnaire information of different customers for each product; according to the questionnaire information and the sales coefficient, analyze the satisfaction of customers with each product, and determine popular products and unpopular products; The sales analysis sub-module: used to obtain the product type of the popular product as the quasi-type; obtain the order data of the quasi-type products sold by different enterprises, and use SARIMA regression to analyze the order quantity differentiation of the popular products in different seasons, and formulate a sales plan for the popular products; A user interaction module: used to summarize the unpopular products and the sales plan and give feedback.

Images