High-anticipation e-commerce platform data processing system and method thereof

By conducting customer group classification and product attribute analysis on e-commerce platform customers, setting push frequency and threshold intervals for different consumer groups, the push frequency management problems of different consumer capabilities in the existing technology are solved, and the transaction rate and push effect are improved.

CN118916764APending Publication Date: 2024-11-08ZHONGZHAO UNITED ELECTRONIC COMMERCE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411072251.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-06
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

It is difficult for existing e-commerce platforms to effectively classify customer groups and push frequency management of customers with different consumption capabilities, resulting in reduced transaction rates and waste of resources.

Method used

Through big data technology, customers' past purchase records are obtained, customer customer groups are classified, and products are pushed according to product attributes and customer group matching degree, push frequency and threshold intervals of different consumer groups are set, and push strategies based on customers and products are realized.

Benefits of technology

The transaction rate is improved, the high-frequency purchase demand of high-consumption customers is ensured, the harassment of low-consumption customers is avoided, resource waste is reduced, and better push effect is achieved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118916764A_ABST
    Figure CN118916764A_ABST
Patent Text Reader

Abstract

The invention discloses a high-anticipation e-commerce platform data processing system and method, and relates to the field. Comprising the following steps: customer data acquisition: acquiring past purchase records of customers through a big data technology; customer group classification: classifying the customers into different customer groups according to previous purchase records; commodity attribute classification: carrying out attribute classification on the commodities according to the purposes of the commodities; and commodity pushing: according to the attribute classification of the commodity and the customer group, selecting the commodity with high matching degree and pushing the commodity to the customer. According to the invention, the customer groups are classified according to the acquired customer data, then the commodity attributes are classified, and pushing is carried out according to different frequencies according to different customer groups and commodity attributes, so that the pre-judgment type pushing not only ensures the high-frequency purchase demand of the high-consumption customer groups, but also ensures the harassment prevention of the low-consumption customer groups, and the user experience is improved. And resource waste is also avoided.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to a field, and in particular to a high-prediction e-commerce platform data processing system and method thereof. Background Art

[0002] With the expansion of online business, more and more e-commerce websites provide users with a large amount of information about goods to be sold. The existing method is that the seller provides all the product information to all consumers. However, with the increase in the number and types of goods for sale, it is difficult for consumers to find their favorite goods in a short time when choosing goods from a large amount of product information, resulting in a lower transaction rate.

[0003] In order to solve the above problems, after searching, the Chinese patent publication number CN105976211A discloses a method and system for pushing product information, including: S1, storing customer tags of M customers, storing product tags and default sorting of N products, and providing an external web page link address and its corresponding product display area. S2, the customer tag of the i-th customer is matched with the product tag of each product to obtain the best sorting of the N products associated with the i-th customer. S3, store the above best sorting in the dynamic buffer, i++, repeat S2 and S3 until the best sorting of the N products associated with the M-th customer is obtained. S4, when receiving an open request for the external web page link address, obtain customer identification information. S5, determine whether there is preset identification information corresponding to the customer identification information. S6, if so, obtain the best sorting of N products corresponding to the preset identification information from the dynamic buffer and send it to the product display area.

[0004] The above patent has the following deficiencies: although it uses customer information to identify customers and then pushes information, it is unable to classify customers due to different spending power and purchase frequency, and thus is unable to determine the corresponding push frequency for customers with different spending power.

[0005] To this end, the present invention proposes a high-prediction e-commerce platform data processing system and method. Summary of the invention

[0006] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a high-prediction e-commerce platform data processing system and method.

[0007] In order to achieve the above object, the present invention adopts the following technical solutions:

[0008] A high-prediction e-commerce platform data processing method comprises the following steps:

[0009] S1: Customer data acquisition, obtaining customers’ previous purchase records through big data technology;

[0010] S2: Customer group classification, classifying customers into different customer groups based on their past purchase records;

[0011] S3: Classification of commodity attributes, classifying commodities according to their uses;

[0012] S4: Product push: According to the product attribute classification and customer group, select highly matching products and push them to customers.

[0013] Preferably: in the step S1, the acquired customer's past purchase records include the customer's name, customer contact information, customer's purchased commodity type, purchase amount and purchase frequency.

[0014] Preferably: customer group classification includes the following steps:

[0015] S21: First, customers are divided into high-consumption, medium-consumption, and low-consumption customer groups based on the amount of goods they purchased and the frequency of purchase;

[0016] S22: Then, secondary classification is performed based on attributes into high-consumption, medium-consumption, and low-consumption customer groups;

[0017] The step S21 specifically includes the following steps:

[0018] S211: Setting the thresholds for high consumption and medium consumptionσ 1 The thresholds of medium and low consumption are 2 ;

[0019] S212: Obtain the total consumption amount based on the customer's previous purchase records Average consumption per time Variance of consumption and consumption frequency i is the number of consumption within 30 days;

[0020] S213: According to the formula Calculate the spending power of your customers and categorize them accordingly.

[0021] Preferably: in S213, a, b, c, d are weight coefficients respectively, which satisfy

[0022] The S213 specifically includes the following situations:

[0023] S2131: If Q ≥ σ 1 , then the customer is a high-spending customer group;

[0024] S2132: If σ 1 >Q>σ 2 , then the customer is a medium-consumption customer group;

[0025] S2133: If Q≤σ 2 , then the customer is a low-spending customer.

[0026] In the step S22, the secondary classification method includes the following steps:

[0027] S221: Obtain items that the user has purchased in the past and classify the items into categories;

[0028] S222: Count the total price H of items of each type;

[0029] S223: Setting multiple classification threshold cutoff points δ 1 ,δ 2 ,......δ n , and the adjacent thresholds and infinite and infinitesimal values ​​are combined into intervals, i.e. (-∞, δ 1 )、......(δ i , δ i+1 )、......(δ n , +∞);

[0030] S224: Calculation By taking the value of and classifying it into the interval, we can get the user classification based on items.

[0031] Preferably: in the step S4, the push method includes a customer-based push method and an item-based push method;

[0032] The client-based push method comprises the following steps:

[0033] S41A: First, select the corresponding push frequencies according to the high-consumption, medium-consumption and low-consumption customer groups, and the push frequencies decrease in descending order from high-consumption, medium-consumption and low-consumption;

[0034] S42A: edit the product information to be pushed, and push the content to three different customer groups according to the corresponding frequency;

[0035] The item-based push method comprises the following steps:

[0036] S41B: Edit the content of the product information to be pushed;

[0037] S42B: Determine the push frequency for each interval based on the ratio of the total price of each type of commodity to the total price of the transaction items;

[0038] S43B: Determine the product type of the product to be pushed, and push it to the customer group matching the type according to the determined push frequency.

[0039] A highly predictive e-commerce platform data processing system comprises a data processing engine, a storage module, an update module, a statistics module and a plan analysis module, wherein the data processing engine is communicatively connected to the storage module, the storage module is used to store customer information, and the customer information is saved to the storage module through registration on a front-end website and entry into a back-end system; the update module is used to update customer attributes, and the staff modifies the customer's attribute information through the back-end; the statistics module comprises a counting module and a scoring module; the plan analysis module is communicatively connected to the storage module, and analyzes and formulates plan tasks based on the data counted by the statistics module.

[0040] Preferably: the customer information includes account information, transaction information and interaction information. The customer account information is the user number, the account password of the corresponding numbered user, the number of operations of the corresponding numbered user in each category template, and the mobile phone number of the corresponding numbered user; the transaction information is the product information of the customer's historical transactions.

[0041] Preferably: the customer information includes account information, transaction information and interaction information. The customer account information is the user number, the account password of the corresponding numbered user, the number of operations of the corresponding numbered user in each category template, and the mobile phone number of the corresponding numbered user; the transaction information is the product information of the customer's historical transactions.

[0042] Preferably: the counting module sorts the customer's order information from few to many or from many to least according to the statistical results of the statistical module, checks the customer's order information for duplication, marks the repeated transaction information, and extracts and sorts the transaction information with the most duplication;

[0043] The data processing engine includes a computing engine and a query engine. The computing engine is used to calculate and analyze data; the query engine is used to query, read and write data.

[0044] Preferably: the query engine includes single-line text search, multi-line text search and cascade search, and the operator performs single-line text search, multi-line text search and cascade search on the page, and cascade search refers to a one-to-many relationship.

[0045] The beneficial effects of the present invention are:

[0046] 1. The present invention classifies customer groups through the acquired customer data, and then classifies product attributes, and then pushes products at different frequencies according to different customer groups and product attributes. This predictive push not only ensures the high-frequency purchase needs of high-spending customer groups, but also prevents harassment of low-spending customer groups and avoids waste of resources.

[0047] 2. The present invention, on the basis of the classification of customer groups and the classification of basic attributes of commodities, adds the threshold interval division of commodity prices, thereby realizing the push based on customers and the push based on commodities, so as to achieve better push effects in different situations. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a flow chart of a high-prediction e-commerce platform data processing system and method proposed by the present invention. DETAILED DESCRIPTION

[0049] The technical solution of the present invention is further described in detail below in conjunction with specific implementation methods.

[0050] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", and "set" should be understood in a broad sense, for example, they can be fixedly connected or set, or detachably connected or set, or integrally connected or set. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0051] Embodiment 1:

[0052] A high-prediction e-commerce platform data processing method comprises the following steps:

[0053] S1: Customer data acquisition, obtaining customers’ previous purchase records through big data technology;

[0054] S2: Customer group classification, classifying customers into different customer groups based on their past purchase records;

[0055] S3: Classification of commodity attributes, classifying commodities according to their uses;

[0056] S4: Product push: According to the product attribute classification and customer group, select highly matching products and push them to customers.

[0057] In the step S1, the customer's past purchase records obtained include the customer's name, customer contact information, the type of goods purchased by the customer, the purchase amount, and the purchase frequency.

[0058] In the step S2, the customer group classification includes the following steps:

[0059] S21: First, customers are divided into high-consumption, medium-consumption, and low-consumption customer groups based on the amount of goods they purchased and the frequency of purchase;

[0060] S22: Then, secondary classification is performed on high-consumption, medium-consumption and low-consumption customer groups based on their attributes.

[0061] The step S21 specifically includes the following steps:

[0062] S211: Setting the thresholds for high consumption and medium consumptionσ 1 The thresholds of medium and low consumption are 2 ;

[0063] S212: Obtain the total consumption amount based on the customer's previous purchase records Average consumption per time Variance of consumption and consumption frequency i is the number of consumption within 30 days;

[0064] S213: According to the formula Calculate the spending power of customers and classify them accordingly

[0065] The S213 specifically includes the following situations:

[0066] S2131: If Q ≥ σ 1 , then the customer is a high-spending customer group;

[0067] S2132: If σ 1 >Q>σ 2 , then the customer is a medium-consumption customer group;

[0068] S2133: If Q≤σ 2 , then the customer is a low-spending customer.

[0069] In the step S22, the secondary classification method includes the following steps:

[0070] S221: Obtain items that the user has purchased in the past and classify the items into categories;

[0071] S222: Count the total price H of items of each type;

[0072] S223: Setting multiple classification threshold cutoff points δ 1 ,δ 2 ,......δ n , and the adjacent thresholds and infinite and infinitesimal values ​​are combined into intervals, i.e. (-∞, δ 1 )、......(δ i , δ i+1 )、......(δ n , +∞);

[0073] S224: Calculation By taking the value of and classifying it into the interval, we can get the user classification based on items.

[0074] In the step S4, the push method is a client-based push method.

[0075] The client-based push method comprises the following steps:

[0076] S41A: First, select the corresponding push frequencies according to the high-consumption, medium-consumption and low-consumption customer groups, and the push frequencies decrease in descending order from high-consumption, medium-consumption and low-consumption;

[0077] S42A: Edit the product information to be pushed, and push the content to three different customer groups at the corresponding frequency.

[0078] Embodiment 2:

[0079] A high-prediction e-commerce platform data processing method comprises the following steps:

[0080] S1: Customer data acquisition, obtaining customers’ previous purchase records through big data technology;

[0081] S2: Customer group classification, classifying customers into different customer groups based on their past purchase records;

[0082] S3: Classification of commodity attributes, classifying commodities according to their uses;

[0083] S4: Product push: According to the product attribute classification and customer group, select highly matching products and push them to customers.

[0084] In the step S1, the customer's past purchase records obtained include the customer's name, customer contact information, the type of goods purchased by the customer, the purchase amount, and the purchase frequency.

[0085] In the step S2, the customer group classification includes the following steps:

[0086] S21: First, customers are divided into high-consumption, medium-consumption, and low-consumption customer groups based on the amount of goods they purchased and the frequency of purchase;

[0087] S22: Then, secondary classification is performed on high-consumption, medium-consumption and low-consumption customer groups based on their attributes.

[0088] The step S21 specifically includes the following steps:

[0089] S211: Setting the thresholds for high consumption and medium consumptionσ 1 The thresholds of medium and low consumption are 2 ;

[0090] S212: Obtain the total consumption amount based on the customer's previous purchase records Average consumption per time Variance of consumption and consumption frequency i is the number of consumption within 30 days;

[0091] S213: According to the formula Calculate the spending power of customers and classify them accordingly

[0092] The S213 specifically includes the following situations:

[0093] S2131: If Q ≥ σ 1 , then the customer is a high-spending customer group;

[0094] S2132: If σ 1 >Q>σ 2 , then the customer is a medium-consumption customer group;

[0095] S2133: If Q≤σ 2 , then the customer is a low-spending customer.

[0096] In the step S22, the secondary classification method includes the following steps:

[0097] S221: Obtain items that the user has purchased in the past and classify the items into categories;

[0098] S222: Count the total price H of items of each type;

[0099] S223: Setting multiple classification threshold cutoff points δ 1 ,δ 2 ,......δ n , and the adjacent thresholds and infinite and infinitesimal values ​​are combined into intervals, i.e. (-∞, δ 1 )、......(δ i , δ i+1 )、......(δ n , +∞);

[0100] S224: Calculation By taking the value of and classifying it into the interval, we can get the user classification based on items.

[0101] In the step S4, the push method is a client-based push method.

[0102] The client-based push method comprises the following steps:

[0103] S41A: First, select the corresponding push frequencies according to the high-consumption, medium-consumption and low-consumption customer groups, and the push frequencies decrease in descending order from high-consumption, medium-consumption and low-consumption;

[0104] S42A: Edit the product information to be pushed, and push the content to three different customer groups at the corresponding frequency.

[0105] Embodiment 3:

[0106] A high-prediction e-commerce platform data processing method comprises the following steps:

[0107] S1: Customer data acquisition, obtaining customers’ previous purchase records through big data technology;

[0108] S2: Customer group classification, classifying customers into different customer groups based on their past purchase records;

[0109] S3: Classification of commodity attributes, classifying commodities according to their uses;

[0110] S4: Product push: According to the product attribute classification and customer group, select highly matching products and push them to customers.

[0111] In the step S1, the customer's past purchase records obtained include the customer's name, customer contact information, the type of goods purchased by the customer, the purchase amount, and the purchase frequency.

[0112] In the step S2, the customer group classification includes the following steps:

[0113] S21: First, customers are divided into high-consumption, medium-consumption, and low-consumption customer groups based on the amount of goods they purchased and the frequency of purchase;

[0114] S22: Then, secondary classification is performed on high-consumption, medium-consumption and low-consumption customer groups based on their attributes.

[0115] The step S21 specifically includes the following steps:

[0116] S211: Setting the thresholds for high consumption and medium consumptionσ 1 The thresholds of medium and low consumption are 2 ;

[0117] S212: Obtain the total consumption amount based on the customer's previous purchase records Average consumption per time Variance of consumption and consumption frequency i is the number of consumption within 30 days;

[0118] S213: According to the formula Calculate the spending power of customers and classify them accordingly

[0119] The S213 specifically includes the following situations:

[0120] S2131: If Q ≥ σ 1 , then the customer is a high-spending customer group;

[0121] S2132: If σ 1 >Q>σ 2 , then the customer is a medium-consumption customer group;

[0122] S2133: If Q≤σ 2 , then the customer is a low-spending customer.

[0123] In the step S22, the secondary classification method includes the following steps:

[0124] S221: Obtain items that the user has purchased in the past and classify the items into categories;

[0125] S222: Count the total price H of items of each type;

[0126] S223: Setting multiple classification threshold cutoff points δ 1 ,δ 2 ,......δ n , and the adjacent thresholds and infinite and infinitesimal values ​​are combined into intervals, i.e. (-∞, δ 1 )、......(δ i , δ i+1 )、......(δ n , +∞);

[0127] S224: Calculation By taking the value of and classifying it into the interval, we can get the user classification based on items.

[0128] In the step S4, the push method is a client-based push method.

[0129] The client-based push method comprises the following steps:

[0130] S41A: First, select the corresponding push frequencies according to the high-consumption, medium-consumption and low-consumption customer groups, and the push frequencies decrease in descending order from high-consumption, medium-consumption and low-consumption;

[0131] S42A: Edit the product information to be pushed, and push the content to three different customer groups at the corresponding frequency.

[0132] Embodiment 4:

[0133] A high-prediction e-commerce platform data processing method comprises the following steps:

[0134] S1: Customer data acquisition, obtaining customers’ previous purchase records through big data technology;

[0135] S2: Customer group classification, classifying customers into different customer groups based on their past purchase records;

[0136] S3: Classification of commodity attributes, classifying commodities according to their uses;

[0137] S4: Product push: According to the product attribute classification and customer group, select highly matching products and push them to customers.

[0138] In the step S1, the customer's past purchase records obtained include the customer's name, customer contact information, the type of goods purchased by the customer, the purchase amount, and the purchase frequency.

[0139] In the step S2, the customer group classification includes the following steps:

[0140] S21: First, customers are divided into high-consumption, medium-consumption, and low-consumption customer groups based on the amount of goods they purchased and the frequency of purchase;

[0141] S22: Then, secondary classification is performed on high-consumption, medium-consumption and low-consumption customer groups based on their attributes.

[0142] The step S21 specifically includes the following steps:

[0143] S211: Setting the thresholds for high consumption and medium consumptionσ 1 The thresholds of medium and low consumption are 2 ;

[0144] S212: Obtain the total consumption amount based on the customer's previous purchase records Average consumption per time Variance of consumption and consumption frequency i is the number of consumption within 30 days;

[0145] S213: According to the formula Calculate the spending power of customers and classify them accordingly

[0146] The S213 specifically includes the following situations:

[0147] S2131: If Q ≥ σ 1 , then the customer is a high-spending customer group;

[0148] S2132: If σ 1 >Q>σ 2 , then the customer is a medium-consumption customer group;

[0149] S2133: If Q≤σ 2 , then the customer is a low-spending customer.

[0150] In the step S22, the secondary classification method includes the following steps:

[0151] S221: Obtain items that the user has purchased in the past and classify the items into categories;

[0152] S222: Count the total price H of items of each type;

[0153] S223: Setting multiple classification threshold cutoff points δ 1 ,δ 2 ,......δ n , and the adjacent thresholds and infinite and infinitesimal values ​​are combined into intervals, i.e. (-∞, δ 1 )、......(δ i , δ i+1 )、......(δ n , +∞);

[0154] S224: Calculation By taking the value of and classifying it into the interval, we can get the user classification based on items.

[0155] In the step S4, the push method is a client-based push method.

[0156] The client-based push method comprises the following steps:

[0157] S41A: First, select the corresponding push frequencies according to the high-consumption, medium-consumption and low-consumption customer groups, and the push frequencies decrease in descending order from high-consumption, medium-consumption and low-consumption;

[0158] S42A: Edit the product information to be pushed, and push the content to three different customer groups at the corresponding frequency.

[0159] Embodiment 5:

[0160] A high-prediction e-commerce platform data processing method comprises the following steps:

[0161] S1: Customer data acquisition, obtaining customers’ previous purchase records through big data technology;

[0162] S2: Customer group classification, classifying customers into different customer groups based on their past purchase records;

[0163] S3: Classification of commodity attributes, classifying commodities according to their uses;

[0164] S4: Product push: According to the product attribute classification and customer group, select highly matching products and push them to customers.

[0165] In the step S1, the customer's past purchase records obtained include the customer's name, customer contact information, the type of goods purchased by the customer, the purchase amount, and the purchase frequency.

[0166] In the step S2, the customer group classification includes the following steps:

[0167] S21: First, customers are divided into high-consumption, medium-consumption, and low-consumption customer groups based on the amount of goods they purchased and the frequency of purchase;

[0168] S22: Then, secondary classification is performed on high-consumption, medium-consumption and low-consumption customer groups based on their attributes.

[0169] The step S21 specifically includes the following steps:

[0170] S211: Setting the thresholds for high consumption and medium consumptionσ 1 The thresholds of medium and low consumption are 2 ;

[0171] S212: Obtain the total consumption amount based on the customer's previous purchase records Average consumption per time Variance of consumption and consumption frequency i is the number of consumption within 30 days;

[0172] S213: According to the formula Calculate the spending power of customers and classify them accordingly

[0173] The S213 specifically includes the following situations:

[0174] S2131: If Q ≥ σ 1 , then the customer is a high-spending customer group;

[0175] S2132: If σ 1 >Q>σ 2 , then the customer is a medium-consumption customer group;

[0176] S2133: If Q≤σ 2 , then the customer is a low-spending customer.

[0177] In the step S22, the secondary classification method includes the following steps:

[0178] S221: Obtain items that the user has purchased in the past and classify the items into categories;

[0179] S222: Count the total price H of items of each type;

[0180] S223: Setting multiple classification threshold cutoff points δ 1 ,δ 2 ,......δ n , and the adjacent thresholds and infinite and infinitesimal values ​​are combined into intervals, i.e. (-∞, δ 1 )、......(δi , δ i+1 )、......(δ n , +∞);

[0181] S224: Calculation By taking the value of and classifying it into the interval, we can get the user classification based on items.

[0182] In the step S4, the push method is a client-based push method.

[0183] The client-based push method comprises the following steps:

[0184] S41A: First, select the corresponding push frequencies according to the high-consumption, medium-consumption and low-consumption customer groups, and the push frequencies decrease in descending order from high-consumption, medium-consumption and low-consumption;

[0185] S42A: Edit the product information to be pushed, and push the content to three different customer groups at the corresponding frequency.

[0186] Embodiment 6:

[0187] A high-prediction e-commerce platform data processing method comprises the following steps:

[0188] S1: Customer data acquisition, obtaining customers’ previous purchase records through big data technology;

[0189] S2: Customer group classification, classifying customers into different customer groups based on their past purchase records;

[0190] S3: Classification of commodity attributes, classifying commodities according to their uses;

[0191] S4: Product push: According to the product attribute classification and customer group, select highly matching products and push them to customers.

[0192] In the step S1, the customer's past purchase records obtained include the customer's name, customer contact information, the type of goods purchased by the customer, the purchase amount, and the purchase frequency.

[0193] In the step S2, the customer group classification includes the following steps:

[0194] S21: First, customers are divided into high-consumption, medium-consumption, and low-consumption customer groups based on the amount of goods they purchased and the frequency of purchase;

[0195] S22: Then, secondary classification is performed on high-consumption, medium-consumption and low-consumption customer groups based on their attributes.

[0196] The step S21 specifically includes the following steps:

[0197] S211: Setting the thresholds for high consumption and medium consumptionσ 1 The thresholds of medium and low consumption are 2 ;

[0198] S212: Obtain the total consumption amount based on the customer's previous purchase records Average consumption per time Variance of consumption and consumption frequency i is the number of consumption within 30 days;

[0199] S213: According to the formula Calculate the spending power of customers and classify them accordingly

[0200] The S213 specifically includes the following situations:

[0201] S2131: If Q ≥ σ 1 , then the customer is a high-spending customer group;

[0202] S2132: If σ 1 >Q>σ 2 , then the customer is a medium-consumption customer group;

[0203] S2133: If Q≤σ 2 , then the customer is a low-spending customer.

[0204] In the step S22, the secondary classification method includes the following steps:

[0205] S221: Obtain items that the user has purchased in the past and classify the items into categories;

[0206] S222: Count the total price H of items of each type;

[0207] S223: Setting multiple classification threshold cutoff points δ 1 ,δ 2 ,......δ n , and the adjacent thresholds and infinite and infinitesimal values ​​are combined into intervals, i.e. (-∞, δ 1 )、......(δ i , δ i+1 )、......(δ n , +∞);

[0208] S224: Calculation By taking the value of and classifying it into the interval, we can get the user classification based on items.

[0209] The item-based push method comprises the following steps:

[0210] S41B: Edit the content of the product information to be pushed;

[0211] S42B: Determine the push frequency for each interval based on the ratio of the total price of each type of commodity to the total price of the transaction items;

[0212] S43B: Determine the product type of the product to be pushed, and push it to the customer group matching the type according to the determined push frequency.

[0213] Embodiment 7:

[0214] A highly predictive e-commerce platform data processing system comprises a data processing engine, a storage module, an update module, a statistics module and a plan analysis module, wherein the data processing engine is communicatively connected to the storage module, the storage module is used to store customer information, and the customer information is saved to the storage module through registration on a front-end website and entry into a back-end system; the update module is used to update customer attributes, and the staff modifies the customer's attribute information through the back-end; the statistics module comprises a counting module and a scoring module; the plan analysis module is communicatively connected to the storage module, and analyzes and formulates plan tasks based on the data counted by the statistics module.

[0215] The customer information includes account information, transaction information and interaction information. The customer account information is the user number, the account password of the corresponding numbered user, the number of operations of the corresponding numbered user on each category template, and the mobile phone number of the corresponding numbered user; the transaction information is the product information of the customer's historical transactions.

[0216] The statistical module is used to count the customer's historical transaction amount, today's transaction amount, this month's transaction amount, the transaction amount in the same period of last month, the refund amount in the past 30 days, and the refund rate in the past 30 days; the statistical module summarizes and counts the customer data stored in the storage module, including statistical analysis of the total amount of various types of transactions of the customer on the platform, and generates statistical reports based on the statistical results;

[0217] The scoring module is used to score customer value and loyalty according to the customer value clustering results. The scoring module rates the customer's capital status, capital liquidity, and transaction information according to the statistical results of the statistical module.

[0218] The counting module sorts the customer's order information from few to many or from many to least according to the statistical results of the statistical module, checks the customer's order information for duplication, marks the repeated transaction information, and extracts and sorts the transaction information with the most duplication;

[0219] The data processing engine includes a computing engine and a query engine. The computing engine is used to calculate and analyze data; the query engine is used to query, read and write data.

[0220] The query engine includes single-line text search, multi-line text search and cascade search. The operator performs single-line text search, multi-line text search and cascade search on the page. Cascade search refers to a one-to-many relationship.

[0221] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A highly predictive e-commerce platform data processing method, It is characterized in that The following steps are involved: S1: Customer data acquisition, obtaining customers’ previous purchase records through big data technology; S2: Customer group classification, classifying customers into different customer groups based on their past purchase records; S3: Classification of commodity attributes, classifying commodities according to their uses; S4: Product push: According to the product attribute classification and customer group, select highly matching products and push them to customers.

2. According to the high-prediction e-commerce platform data processing method of claim 1, It is characterized in that In the step S1, the customer's past purchase records obtained include the customer's name, customer contact information, the type of goods purchased by the customer, the purchase amount, and the purchase frequency.

3. According to the high-prediction e-commerce platform data processing method of claim 1, It is characterized in that In the step S2, the customer group classification includes the following steps: S21: First, customers are divided into high-consumption, medium-consumption, and low-consumption customer groups based on the amount of goods they purchased and the frequency of purchase; S22: Then, secondary classification is performed based on attributes into high-consumption, medium-consumption, and low-consumption customer groups; The step S21 specifically includes the following steps: S211: Setting the thresholds for high consumption and medium consumptionσ 1 The thresholds of medium and low consumption are 2 ; S212: Obtain the total consumption amount based on the customer's previous purchase records Average consumption per time Variance of consumption and consumption frequency i is the number of consumption within 30 days; S213: According to the formula Calculate the spending power of your customers and categorize them accordingly.

4. According to the high-prediction e-commerce platform data processing method of claim 3, It is characterized in that In S213, a, b, c, and d are weight coefficients respectively, which satisfy The S213 specifically includes the following situations: S2131: If Q ≥ σ 1 , then the customer is a high-spending customer group; S2132: If σ 1 >Q>σ 2 , then the customer is a medium-consumption customer group; S2133: If Q≤σ 2 , then the customer is a low-spending customer. In the step S22, the secondary classification method includes the following steps: S221: Obtain items that the user has purchased in the past and classify the items into categories; S222: Count the total price H of items of each type; S223: Setting multiple classification threshold cutoff points δ 1 ,δ 2 ,......δ n , and the adjacent thresholds and infinite and infinitesimal values ​​are combined into intervals, i.e. (-∞, δ 1 )、......(δ i , δ i+1 )、......(δ n , +∞); S224: Calculation By taking the value of and classifying it into the interval, we can get the user classification based on items.

5. According to the high-prediction e-commerce platform data processing method of claim 1, It is characterized in that In the step S4, the push method includes a customer-based push method and an item-based push method; The client-based push method comprises the following steps: S41A: First, select the corresponding push frequencies according to the high-consumption, medium-consumption and low-consumption customer groups, and the push frequencies decrease in descending order from high-consumption, medium-consumption and low-consumption; S42A: edit the product information to be pushed, and push the content to three different customer groups according to the corresponding frequency; The item-based push method comprises the following steps: S41B: Edit the content of the product information to be pushed; S42B: Determine the push frequency for each interval based on the ratio of the total price of each type of commodity to the total price of the transaction items; S43B: Determine the product type of the product to be pushed, and push it to the customer group matching the type according to the determined push frequency.

6. A high-prediction e-commerce platform data processing system, which is used to implement the high-prediction e-commerce platform data processing method described in any one of claims 1-5, It is characterized in that It includes a data processing engine, a storage module, an update module, a statistics module and a plan analysis module, and is characterized in that the data processing engine is communicatively connected to the storage module, the storage module is used to store customer information, and the customer information is saved to the storage module through registration on the front-end website and entry into the back-end system; the update module is used to update customer attributes, and the staff modifies the customer's attribute information through the back-end; the statistics module includes a counting module and a scoring module; the plan analysis module is communicatively connected to the storage module, and analyzes and formulates planned tasks based on the data counted by the statistics module.

7. According to the high-prediction e-commerce platform data processing system of claim 6, It is characterized in that The customer information includes account information, transaction information and interaction information. The customer account information is the user number, the account password of the corresponding numbered user, the number of operations of the corresponding numbered user on each category template, and the mobile phone number of the corresponding numbered user; the transaction information is the product information of the customer's historical transactions.

8. According to the high-prediction e-commerce platform data processing system of claim 6, It is characterized in that The customer information includes account information, transaction information and interaction information. The customer account information is the user number, the account password of the corresponding numbered user, the number of operations of the corresponding numbered user on each category template, and the mobile phone number of the corresponding numbered user; the transaction information is the product information of the customer's historical transactions.

9. According to the high-prediction e-commerce platform data processing system of claim 6, It is characterized in that The counting module sorts the customer's order information from few to many or from many to least according to the statistical results of the statistical module, checks the customer's order information for duplication, marks the repeated transaction information, and extracts and sorts the transaction information with the most duplication; The data processing engine includes a computing engine and a query engine. The computing engine is used to calculate and analyze data; the query engine is used to query, read and write data.

10. The high-prediction e-commerce platform data processing system according to claim 6, It is characterized in that The query engine includes single-line text search, multi-line text search and cascade search. The operator performs single-line text search, multi-line text search and cascade search on the page. Cascade search refers to a one-to-many relationship.