E-commerce operation management system based on big data

Through the e-commerce operation and management system based on big data, we collect and analyze customer behavior, generate keywords and correlation numbers, and push related products, and solve the problem of how to capture user needs and enhance user stickiness in e-commerce, and realize personalized shopping Improved experience and user stickiness.

CN119027173AInactive Publication Date: 2024-11-26陆孝明
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202411110419.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-14
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

In e-commerce, it is difficult to effectively realize a personalized shopping experience in existing technology, how to capture user needs, expand user shopping tendencies, and enhance user stickiness.

Method used

Design an e-commerce operation management system based on big data, including control center, database, data acquisition module, data processing module, operation strategy generation module and early warning module. By collecting customer behavior, generate keywords, effective prices and policy update cycles , calculate the internal correlation contact number and external correlation contact number, send push information to customers regularly, process early warning information, capture user needs and push related products.

Benefits of technology

By obtaining customer shopping keywords, clarifying user needs, pushing products related to keywords, expanding the range of optional users' shopping, increasing platform interaction, and improving user stickiness.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119027173A_ABST
    Figure CN119027173A_ABST
Patent Text Reader

Abstract

The invention discloses an e-commerce operation management system based on big data, and relates to the technical field of big data, the e-commerce operation management system comprises a control center, the control center is connected with a database, a data acquisition module, a data processing module, an operation strategy generation module and an early warning module; the database is used for storing commodity information, customer information and shop information; the data acquisition module is used for acquiring customer behaviors, setting an acquisition period according to the customer behaviors, and generating corresponding early warning information by acquiring the customer behaviors in the acquisition period; the data processing module is used for generating a keyword, an effective price and a strategy updating period of a customer according to customer behaviors, and generating an internal correlation coefficient and an external correlation coefficient according to commodity information and shop information; and the operation strategy generation module is used for generating corresponding push information according to the keyword, the valid price, the internal correlation coefficient and the external correlation coefficient, and regularly sending the push information to the customer according to the strategy update period of the customer.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of big data technology, and in particular to an e-commerce operation management system based on big data. Background Art

[0002] E-commerce is a business activity centered on the exchange of goods using information network technology. People are not restricted by time, space, or the many restrictions of traditional shopping, and can trade online anytime and anywhere. Nowadays, people are increasingly pursuing fashion, individuality, and the shopping environment. Online shopping can better reflect the personalized shopping process.

[0003] As people pursue personalized shopping through e-commerce malls, how to capture user needs, expand user shopping tendencies, and enhance user stickiness is the issue we currently need to consider. Summary of the invention

[0004] In order to solve the above problems, the purpose of the present invention is to provide an e-commerce operation management system based on big data.

[0005] The object of the present invention can be achieved by the following technical solutions: An e-commerce operation management system based on big data, comprising a control center, wherein the control center is communicatively connected to a database, a data acquisition module, a data processing module, an operation strategy generation module and an early warning module;

[0006] The database is used to store product information, customer information and store information;

[0007] The data collection module is used to collect customer behavior, set a collection period according to the customer behavior, and generate corresponding warning information by acquiring the customer behavior within the collection period;

[0008] The data processing module is used to generate the customer's keywords, effective prices and strategy update cycle according to the customer's behavior, and to generate internal correlation coefficients and external correlation coefficients according to product information and store information;

[0009] The operation strategy generation module is used to generate corresponding push information according to keywords, effective prices, internal correlation coefficients and external correlation coefficients, and regularly send push information to the customer according to the customer's strategy update cycle;

[0010] The warning module is used to process warning information;

[0011] Furthermore, the database is used to store product information, customer information and store information;

[0012] The database includes commodity units, customer units and store units;

[0013] The commodity unit is used to store commodity information, which includes commodity ID, commodity name, commodity major category, commodity minor category, commodity price and commodity link; wherein the commodity major categories include clothing, food, housing, transportation and others, and the commodity minor category is a subdivision of the commodity, which is marked by the user;

[0014] The customer unit is used to store customer information, which includes customer ID, customer name, customer gender, contact information, customer behavior, behavior occurrence time, behavior occurrence duration and keywords;

[0015] The shop unit is used to store shop information, which includes shop ID, shop name and commodity group; wherein the commodity group is composed of a plurality of commodity links.

[0016] Furthermore, the process of collecting customer behaviors, setting a collection period according to the customer behaviors, and generating corresponding warning information by obtaining customer behaviors within the collection period includes:

[0017] The customer behavior includes the behavior type and the behavior object;

[0018] The behavior types include browsing behavior, purchasing behavior and return behavior;

[0019] The behavior object includes a product link;

[0020] Set up the acquisition unit and monitoring unit;

[0021] The collection unit is used to collect customer behavior;

[0022] When the collection unit collects customer behavior of the return behavior, it obtains the customer ID, generates a monitoring instruction according to the customer ID, and submits the monitoring instruction to the monitoring unit, which then processes the instruction accordingly;

[0023] The monitoring unit is used to receive monitoring instructions and set a collection cycle, obtain customer behavior within the customer collection cycle, generate a cycle return rate based on the customer behavior, and generate corresponding warning information based on the cycle return rate.

[0024] Furthermore, the process of collecting customer behavior includes:

[0025] Obtain customer behavior type and behavior object, and generate customer behavior;

[0026] If a customer browses product A on an e-commerce platform, a product link A corresponding to product A is obtained, and a customer behavior is generated according to the product link A; the behavior type corresponding to the customer behavior is a browsing behavior, and the behavior object corresponding to the customer behavior is the product link A;

[0027] If a customer purchases product B on an e-commerce platform, a product link B corresponding to product B is obtained, and a customer behavior is generated according to the product link B; the behavior type corresponding to the customer behavior is a purchase behavior, and the behavior object corresponding to the customer behavior is product link B;

[0028] If the customer returns product C on the e-commerce platform, the product link C corresponding to product C is obtained, and a customer behavior is generated according to the product link C; the behavior type corresponding to the customer behavior is a return behavior, and the behavior object corresponding to the customer behavior is the product link C;

[0029] Obtain the collection time of the customer behavior, mark the collection time as the behavior occurrence time, and perform timing to obtain the duration of the behavior;

[0030] The customer ID corresponding to the customer whose customer behavior is collected is obtained, and based on the customer ID, the customer behavior, the time when the behavior occurred, and the duration of the behavior are sent to the database for storage.

[0031] Furthermore, the process of obtaining customer behavior within a customer acquisition cycle, generating a periodic return rate according to the customer behavior, and generating corresponding warning information according to the periodic return rate includes:

[0032] Obtain the quantity corresponding to each behavior type in the customer behavior;

[0033] The number of customer behaviors whose behavior type is browsing behavior is recorded as browsing value, the number of customer behaviors whose behavior type is purchasing behavior is recorded as purchasing value, and the number of customer behaviors whose behavior type is returning behavior is recorded as returning value;

[0034] Generate cycle return rate based on browsing value, purchase value and return value;

[0035] Said

[0036] Set return thresholds;

[0037] If the period return rate is greater than or equal to the return threshold, a first-level warning message is generated based on the customer ID and sent to the warning module, which will take corresponding action; otherwise, no action is required.

[0038] Furthermore, the process of generating keywords, effective prices and strategy update cycles for the customer based on the customer behavior includes:

[0039] The keywords include strong keywords and weak keywords;

[0040] The effective price includes an effective price anchor point and an effective price range;

[0041] Set the behavior update cycle and set the effective time domain according to the behavior update cycle;

[0042] Acquire customer behaviors within a valid time domain in a database, mark the customer behaviors as valid behaviors, acquire keywords based on the valid behaviors, and send the keywords to a database for storage;

[0043] Set the duration threshold;

[0044] Obtain the duration of the behavior corresponding to the valid behavior whose behavior type is browsing behavior. If the duration of the behavior is less than the duration threshold, mark the valid behavior as invalid behavior. Otherwise, no operation is required.

[0045] Obtaining a behavior type from valid behaviors, obtaining a behavior object corresponding to the behavior type according to the behavior type, and then obtaining a commodity name, commodity major category, and commodity minor category corresponding to the behavior object, recording them as vocabulary, and generating a behavior-related vocabulary set;

[0046] Set display threshold;

[0047] According to the behavior-related vocabulary set, the proportion of each word is obtained. If the proportion is greater than or equal to the display threshold, the word is marked as a behavior keyword; if the proportion is less than the display threshold, all words are marked as behavior keywords;

[0048] Mark the behavior keywords corresponding to the behavior types of purchase behavior and browsing behavior as strong keywords, and mark the behavior keywords corresponding to the behavior type of return behavior as weak keywords;

[0049] If there are overlapping words between strong keywords and weak keywords, delete the overlapping words;

[0050] Get keywords and send them to the database for storage;

[0051] According to the effective behavior, obtain the product price corresponding to the strong keyword. If only one product price is obtained, mark the product price as a valid price anchor point. If more than one product price is obtained, generate a valid price range based on the product price.

[0052] Obtaining the behavior occurrence time corresponding to the valid behavior whose behavior type is purchase behavior, and marking the occurrence time as the valid purchase time;

[0053] Obtain the behavior occurrence time corresponding to the valid behavior whose behavior type is return behavior, and mark the occurrence time as the valid return time;

[0054] Establish a one-dimensional timeline and map the effective purchase time and the effective return time to the one-dimensional timeline;

[0055] Set unit time domain and behavior threshold;

[0056] If there are a valid purchase times in the time range of the unit time domain mapping on the one-dimensional time axis, then the time range is marked as a valid purchase time zone; where a is greater than the behavior threshold;

[0057] If there are b valid return times within the time range of the unit time domain mapping on the one-dimensional time axis, then the time range is marked as a valid return time zone; where b is greater than the behavior threshold;

[0058] Delete the time zones that overlap with the valid return time zones in the valid purchase time zones, and generate valid push time zones;

[0059] Obtain the start time of each valid push time zone, and generate several time intervals according to the start time;

[0060] Get the mean, mode and median corresponding to several time intervals;

[0061] Generate a strategy update cycle based on the mean, mode and median;

[0062] Said

[0063] Furthermore, the process of generating the internal correlation coefficient and the external correlation coefficient according to the product information and the store information includes:

[0064] Randomly obtain a product and mark the product as a target product;

[0065] Obtain the internal correlation coefficient and external correlation coefficient between the target product and other products;

[0066] The other commodities include internal commodities and external commodities;

[0067] Obtain a product link corresponding to a target product, and mark the product link as a target link;

[0068] Obtain a product link in the same product group as the target link, and mark the product ID corresponding to the product link as an internal ID;

[0069] Mark the product corresponding to the internal ID as an internal product;

[0070] Mark the products corresponding to the product IDs other than the internal ID in the product unit as external products;

[0071] Set correlation coefficient_A, correlation coefficient_B and correlation coefficient_C;

[0072] Obtain the product name, major category, minor category and price of the target product, mark them as name α, major category α, minor category α and price α respectively, and record them as target product information;

[0073] Get the product name, major category, minor category and price of the internal product, mark them as name β, major category β, minor category β and price β respectively, and record them as internal product information;

[0074] Obtain the product name, major category, minor category and price corresponding to the external product, mark them as name γ, major category γ, minor category γ and price γ respectively, and record them as external product information;

[0075] According to the target product information, the internal product information, the correlation coefficient_A, the correlation coefficient_B and the correlation coefficient_C, the internal correlation coefficient and the external correlation coefficient between the internal product and the target product are obtained;

[0076] The internal correlation coefficient = SIMILARTO (correlation coefficient_C, name α, name β) + SIMILARTO (correlation coefficient_A, large category α, large category β) + SIMILARTO (correlation coefficient_A, small category α, small category β) + MULWITH (correlation coefficient_C, price α, price β);

[0077] The external correlation coefficient = SIMILARTO (correlation coefficient_C, name α, name β) + SIMILARTO (correlation coefficient_B, major category α, major category β) + MULWITH (correlation coefficient_C, price α, price β);

[0078] Among them, SIMILARTO (x, y, z) = x × similarTo (y, z), MULWITH (x, y, z) = x × mulWith (y, z);

[0079] When i and j are the same in the similarTo(i, j) function, it returns 1, otherwise it returns 0;

[0080] mulWith(t, p) function returns |

[0081] According to the target product information, the internal product information, the external product information, the correlation coefficient_B and the correlation coefficient_C, the internal correlation coefficient and the external correlation coefficient between the external product and the target product are obtained;

[0082] The internal correlation coefficient = SIMILAR_name + SIMILAR_large category + SIMILAR_small category + MUL_price;

[0083] in,

[0084] The external correlation coefficient = SIMILARTO (correlation coefficient_C, name α, name γ) + SIMILARTO (correlation coefficient_B, large category α, large category γ) + SIMILARTO (correlation coefficient_B, small category α, small category γ) + MULWITH (correlation coefficient_C, price α, price γ).

[0085] Furthermore, the process of generating corresponding push information according to the keywords, effective prices, internal correlation coefficients and external correlation coefficients includes:

[0086] Set the positive effective percentage and negative effective percentage, where the positive effective percentage is greater than the negative effective percentage;

[0087] Generate push price range based on positive effective percentage, negative effective percentage and effective price;

[0088] If the effective price is only one effective price anchor point, the pushed price range = [effective price anchor point × negative effective percentage, effective price anchor point × positive effective percentage];

[0089] If the valid price has only one valid price range, then the endpoint values ​​of the valid price range are obtained, recorded as the left end value and the right end value, wherein the left end value is less than the right end value, and the push price range is generated according to the left end value, the right end value, the positive effective percentage and the negative effective percentage, and the push price range = [left end value × negative effective percentage, right end value × positive effective percentage];

[0090] Otherwise, according to the principle of generating a push price range from a valid price anchor point and the principle of generating a push price range from a valid price range, a corresponding push price partition is generated through the valid price and the valid percentage, a union between the push price partitions is obtained, and a push price range is generated according to the union;

[0091] Set targeting thresholds and push volume values;

[0092] Obtain the product name, product category and product subcategory in the product information and record them as information sub-items; if there are num products in the three information sub-items that are consistent with the strong keyword, mark the product as a targeted product; where num is greater than or equal to the targeting threshold;

[0093] Obtain the internal and external products corresponding to the targeted product;

[0094] Obtaining an internal correlation coefficient and an external correlation coefficient between the targeted product and the internal product, and generating a push value for the internal product according to the internal correlation coefficient and the external correlation coefficient corresponding to the internal product;

[0095] Obtaining an internal correlation coefficient and an external correlation coefficient between the target product and the external product, and generating a push value for the external product according to the internal correlation coefficient and the external correlation coefficient corresponding to the external product;

[0096] The push value = internal correlation coefficient + external correlation coefficient;

[0097] According to the push value, internal and external products are arranged separately;

[0098] If there are several commodities that cannot be arranged due to equal push values, they will be arranged according to the commodity prices; if the commodity prices are equal, the commodities will be arranged randomly;

[0099] After the internal and external products are arranged, the first pushed products whose prices are within the pushed price range are obtained, the products are marked as pushed products, and push information is generated according to the pushed products; wherein push is equal to the push amount value.

[0100] Furthermore, the process of processing the warning information includes:

[0101] If a first-level warning message is received, query information is generated based on the customer ID, and the query information is submitted to the database to obtain the contact information corresponding to the customer ID. A reminder message is sent to the customer based on the contact information, informing the customer to purchase with caution.

[0102] Compared with the prior art, the present invention has the following beneficial effects:

[0103] 1. Clarify customer needs by obtaining the keywords that customers use to shop;

[0104] 2. Obtain targeted products based on keywords, and obtain the internal and external correlation coefficients between the targeted products and other products, and use the correlation coefficients to explain the correlation and extensibility between other products and the targeted products;

[0105] 3. Capture extended products based on customer keywords through internal and external correlation coefficients. Do not stick to pushing only products that fully meet customer keyword requirements. Push products that have a certain relevance to keyword requirements to customers, thereby expanding customers' shopping options, increasing interactivity with the platform, and improving user stickiness. BRIEF DESCRIPTION OF THE DRAWINGS

[0106] Figure 1 It is a schematic diagram of the present invention. DETAILED DESCRIPTION

[0107] like Figure 1As shown, an e-commerce operation management system based on big data includes a control center, and the control center is connected to a database, a data acquisition module, a data processing module, an operation strategy generation module and an early warning module;

[0108] The database is used to store product information, customer information and store information;

[0109] The data collection module is used to collect customer behavior, set a collection period according to the customer behavior, and generate corresponding warning information by acquiring the customer behavior within the collection period;

[0110] The data processing module is used to generate the customer's keywords, effective prices and strategy update cycle according to the customer's behavior, and to generate internal correlation coefficients and external correlation coefficients according to product information and store information;

[0111] The operation strategy generation module is used to generate corresponding push information according to keywords, effective prices, internal correlation coefficients and external correlation coefficients, and regularly send push information to the customer according to the customer's strategy update cycle;

[0112] The warning module is used to process warning information;

[0113] It should be further explained that, in the specific implementation process, the database is used to store product information, customer information and store information;

[0114] The database includes commodity units, customer units and store units;

[0115] The commodity unit is used to store commodity information, which includes commodity ID, commodity name, commodity major category, commodity minor category, commodity price and commodity link; wherein the commodity major categories include clothing, food, housing, transportation and others, and the commodity minor category is a subdivision of the commodity, which is marked by the user;

[0116] The customer unit is used to store customer information, which includes customer ID, customer name, customer gender, contact information, customer behavior, behavior occurrence time, behavior occurrence duration and keywords;

[0117] The shop unit is used to store shop information, which includes shop ID, shop name and product group; wherein the product group is composed of a plurality of product links;

[0118] It should be further explained that, in the specific implementation process, the process of collecting customer behaviors, setting a collection cycle according to the customer behaviors, and generating corresponding warning information by obtaining customer behaviors within the collection cycle includes:

[0119] The customer behavior includes the behavior type and the behavior object;

[0120] The behavior types include browsing behavior, purchasing behavior and return behavior;

[0121] The behavior object includes a product link;

[0122] Set up the acquisition unit and monitoring unit;

[0123] The collection unit is used to collect customer behavior;

[0124] When the collection unit collects customer behavior of the return behavior, it obtains the customer ID, generates a monitoring instruction according to the customer ID, and submits the monitoring instruction to the monitoring unit, which then processes the instruction accordingly;

[0125] The monitoring unit is used to receive monitoring instructions and set a collection cycle, obtain customer behavior within the customer collection cycle, generate a cycle return rate according to the customer behavior, and generate corresponding warning information according to the cycle return rate;

[0126] The process of collecting customer behavior includes:

[0127] Obtain customer behavior type and behavior object, and generate customer behavior;

[0128] If a customer browses product A on an e-commerce platform, a product link A corresponding to product A is obtained, and a customer behavior is generated according to the product link A; the behavior type corresponding to the customer behavior is a browsing behavior, and the behavior object corresponding to the customer behavior is the product link A;

[0129] If a customer purchases product B on an e-commerce platform, a product link B corresponding to product B is obtained, and a customer behavior is generated according to the product link B; the behavior type corresponding to the customer behavior is a purchase behavior, and the behavior object corresponding to the customer behavior is product link B;

[0130] If the customer returns product C on the e-commerce platform, the product link C corresponding to product C is obtained, and a customer behavior is generated according to the product link C; the behavior type corresponding to the customer behavior is a return behavior, and the behavior object corresponding to the customer behavior is the product link C;

[0131] Obtain the collection time of the customer behavior, mark the collection time as the behavior occurrence time, and perform timing to obtain the duration of the behavior;

[0132] Obtain the customer ID corresponding to the customer whose customer behavior is collected, and send the customer behavior, the time when the behavior occurred, and the duration of the behavior to the database for storage based on the customer ID;

[0133] The process of obtaining customer behavior within a customer collection period, generating a period return rate according to the customer behavior, and generating corresponding warning information according to the period return rate includes:

[0134] Obtain the quantity corresponding to each behavior type in the customer behavior;

[0135] The number of customer behaviors whose behavior type is browsing behavior is recorded as browsing value, the number of customer behaviors whose behavior type is purchasing behavior is recorded as purchasing value, and the number of customer behaviors whose behavior type is returning behavior is recorded as returning value;

[0136] Generate cycle return rate based on browsing value, purchase value and return value;

[0137] Said

[0138] Set return thresholds;

[0139] If the period return rate is greater than or equal to the return threshold, a first-level warning message is generated based on the customer ID and sent to the warning module, which will take corresponding actions. Otherwise, no action is required.

[0140] It should be further explained that, in the specific implementation process, the process of generating the customer's keywords, effective prices and strategy update cycle based on the customer's behavior includes:

[0141] The keywords include strong keywords and weak keywords;

[0142] The effective price includes an effective price anchor point and an effective price range;

[0143] Set the behavior update cycle and set the effective time domain according to the behavior update cycle;

[0144] Acquire customer behaviors within a valid time domain in a database, mark the customer behaviors as valid behaviors, acquire keywords based on the valid behaviors, and send the keywords to a database for storage;

[0145] Set duration threshold;

[0146] Obtain the duration of the behavior corresponding to the valid behavior whose behavior type is browsing behavior. If the duration of the behavior is less than the duration threshold, mark the valid behavior as invalid behavior. Otherwise, no operation is required.

[0147] Obtaining a behavior type from valid behaviors, obtaining a behavior object corresponding to the behavior type according to the behavior type, and then obtaining a commodity name, commodity major category, and commodity minor category corresponding to the behavior object, recording them as vocabulary, and generating a behavior-related vocabulary set;

[0148] Set display threshold;

[0149] According to the behavior-related vocabulary set, the proportion of each word is obtained. If the proportion is greater than or equal to the display threshold, the word is marked as a behavior keyword; if the proportion is less than the display threshold, all words are marked as behavior keywords;

[0150] Mark the behavior keywords corresponding to the behavior types of purchase behavior and browsing behavior as strong keywords, and mark the behavior keywords corresponding to the behavior type of return behavior as weak keywords;

[0151] If there are overlapping words between strong keywords and weak keywords, delete the overlapping words;

[0152] Get keywords and send them to the database for storage;

[0153] According to the effective behavior, obtain the product price corresponding to the strong keyword. If only one product price is obtained, mark the product price as a valid price anchor point. If more than one product price is obtained, generate a valid price range based on the product price.

[0154] Obtaining the behavior occurrence time corresponding to the valid behavior whose behavior type is purchase behavior, and marking the occurrence time as the valid purchase time;

[0155] Obtain the behavior occurrence time corresponding to the valid behavior whose behavior type is return behavior, and mark the occurrence time as the valid return time;

[0156] Establish a one-dimensional timeline and map the effective purchase time and the effective return time to the one-dimensional timeline;

[0157] Set unit time domain and behavior threshold;

[0158] If there are a valid purchase times in the time range of the unit time domain mapping on the one-dimensional time axis, then the time range is marked as a valid purchase time zone; where a is greater than the behavior threshold;

[0159] If there are b valid return times within the time range of the unit time domain mapping on the one-dimensional time axis, then the time range is marked as a valid return time zone; where b is greater than the behavior threshold;

[0160] Delete the time zones that overlap with the valid return time zones in the valid purchase time zones, and generate valid push time zones;

[0161] Obtain the start time of each valid push time zone, and generate several time intervals according to the start time;

[0162] Get the mean, mode and median corresponding to several time intervals;

[0163] Generate a strategy update cycle based on the mean, mode and median;

[0164] Said

[0165] It should be further explained that, in the specific implementation process, the process of generating the internal correlation coefficient and the external correlation coefficient according to the product information and the store information includes:

[0166] Randomly obtain a product and mark the product as a target product;

[0167] Obtain the internal correlation coefficient and external correlation coefficient between the target product and other products;

[0168] The other commodities include internal commodities and external commodities;

[0169] Obtain a product link corresponding to a target product, and mark the product link as a target link;

[0170] Obtain a product link in the same product group as the target link, and mark the product ID corresponding to the product link as an internal ID;

[0171] Mark the product corresponding to the internal ID as an internal product;

[0172] Mark the products corresponding to the product IDs other than the internal ID in the product unit as external products;

[0173] Set correlation coefficient_A, correlation coefficient_B and correlation coefficient_C;

[0174] Obtain the product name, major category, minor category and price of the target product, mark them as name α, major category α, minor category α and price α respectively, and record them as target product information;

[0175] Get the product name, major category, minor category and price of the internal product, mark them as name β, major category β, minor category β and price β respectively, and record them as internal product information;

[0176] Obtain the product name, major category, minor category and price corresponding to the external product, mark them as name γ, major category γ, minor category γ and price γ respectively, and record them as external product information;

[0177] According to the target product information, the internal product information, the correlation coefficient_A, the correlation coefficient_B and the correlation coefficient_C, the internal correlation coefficient and the external correlation coefficient between the internal product and the target product are obtained;

[0178] The internal correlation coefficient = SIMILARTO (correlation coefficient_C, name α, name β) + SIMILARTO (correlation coefficient_A, large category α, large category β) + SIMILARTO (correlation coefficient_A, small category α, small category β) + MULWITH (correlation coefficient_C, price α, price β);

[0179] The external correlation coefficient = SIMILARTO (correlation coefficient_C, name α, name β) + SIMILARTO (correlation coefficient_B, major category α, major category β) + MULWITH (correlation coefficient_C, price α, price β);

[0180] Among them, SIMILARTO(x, y, z) = x × similarTo (y, z), MULWITH (x, y, z) = x × mulWith (y, z);

[0181] When i and j are the same in the similarTo(i, j) function, it returns 1, otherwise it returns 0;

[0182] mulWith(t, p) function returns

[0183] According to the target product information, the internal product information, the external product information, the correlation coefficient_B and the correlation coefficient_C, the internal correlation coefficient and the external correlation coefficient between the external product and the target product are obtained;

[0184] The internal correlation coefficient = SIMILAR_name + SIMILAR_large category + SIMILAR_small category + MUL_price;

[0185] in,

[0186] The external correlation coefficient = SIMILARTO (correlation coefficient_C, name α, name γ) + SIMILARTO (correlation coefficient_B, large category α, large category γ) + SIMILARTO (correlation coefficient_B, small category α, small category γ) + MULWITH (correlation coefficient_C, price α, price γ);

[0187] It should be further explained that, in the specific implementation process, the process of generating corresponding push information according to keywords, effective prices, internal correlation coefficients and external correlation coefficients includes:

[0188] Set the positive effective percentage and negative effective percentage, where the positive effective percentage is greater than the negative effective percentage;

[0189] Generate push price range based on positive effective percentage, negative effective percentage and effective price;

[0190] If the effective price is only one effective price anchor point, the pushed price range = [effective price anchor point × negative effective percentage, effective price anchor point × positive effective percentage];

[0191] If the valid price has only one valid price range, then the endpoint values ​​of the valid price range are obtained, recorded as the left end value and the right end value, wherein the left end value is less than the right end value, and the push price range is generated according to the left end value, the right end value, the positive effective percentage and the negative effective percentage, and the push price range = [left end value × negative effective percentage, right end value × positive effective percentage];

[0192] Otherwise, according to the principle of generating a push price range from a valid price anchor point and the principle of generating a push price range from a valid price range, a corresponding push price partition is generated through the valid price and the valid percentage, a union between the push price partitions is obtained, and a push price range is generated according to the union;

[0193] Set targeting thresholds and push volume values;

[0194] Obtain the product name, product category and product subcategory in the product information and record them as information sub-items; if there are num products in the three information sub-items that are consistent with the strong keyword, mark the product as a targeted product; where num is greater than or equal to the targeting threshold;

[0195] Obtain the internal and external products corresponding to the targeted product;

[0196] Obtaining an internal correlation coefficient and an external correlation coefficient between the targeted product and the internal product, and generating a push value for the internal product according to the internal correlation coefficient and the external correlation coefficient corresponding to the internal product;

[0197] Obtaining an internal correlation coefficient and an external correlation coefficient between the target product and the external product, and generating a push value for the external product according to the internal correlation coefficient and the external correlation coefficient corresponding to the external product;

[0198] The push value = internal correlation coefficient + external correlation coefficient;

[0199] According to the push value, internal and external products are arranged separately;

[0200] If there are several commodities that cannot be arranged due to equal push values, they will be arranged according to the commodity prices; if the commodity prices are equal, the commodities will be arranged randomly;

[0201] Get the first pushed products whose prices are within the push price range after the internal and external products are arranged, mark the products as pushed products, and generate push information based on the pushed products; where push is equal to the push amount value;

[0202] It should be further explained that, in the specific implementation process, the process of handling early warning information includes:

[0203] If a first-level warning message is received, query information is generated based on the customer ID, the query information is submitted to the database, the contact information corresponding to the customer ID is obtained, and a reminder message is sent to the customer based on the contact information, informing the customer to purchase with caution;

[0204] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An e-commerce operation management system based on big data, including a control center, characterized in that: The control center is communicatively connected to a database, a data acquisition module, a data processing module, an operation strategy generation module and an early warning module; The database is used to store product information, customer information and store information; The data collection module is used to collect customer behavior, set a collection period according to the customer behavior, and generate corresponding warning information by acquiring the customer behavior within the collection period; The data processing module is used to generate the customer's keywords, effective prices and strategy update cycle according to the customer's behavior, and to generate internal correlation coefficients and external correlation coefficients according to product information and store information; The operation strategy generation module is used to generate corresponding push information according to keywords, effective prices, internal correlation coefficients and external correlation coefficients, and regularly send push information to the customer according to the customer's strategy update cycle; The early warning module is used to process early warning information.

2. According to the big data-based e-commerce operation management system of claim 1, it is characterized in that: The database is used to store product information, customer information and store information; The database includes commodity units, customer units and store units; The commodity unit is used to store commodity information, which includes commodity ID, commodity name, commodity major category, commodity minor category, commodity price and commodity link; wherein the commodity major categories include clothing, food, housing, transportation and others, and the commodity minor category is a subdivision of the commodity, which is marked by the user; The customer unit is used to store customer information, which includes customer ID, customer name, customer gender, contact information, customer behavior, behavior occurrence time, behavior occurrence duration and keywords; The shop unit is used to store shop information, which includes shop ID, shop name and commodity group; wherein the commodity group is composed of a plurality of commodity links.

3. The e-commerce operation management system based on big data according to claim 2 is characterized in that: The process of collecting customer behaviors, setting a collection cycle based on customer behaviors, and generating corresponding warning information by obtaining customer behaviors within the collection cycle includes: The customer behavior includes a behavior type and a behavior object. The behavior type includes browsing behavior, purchasing behavior and return behavior. The behavior object includes a product link. Set up the acquisition unit and monitoring unit; The collection unit is used to collect customer behavior; When the collection unit collects customer behavior of the return behavior, it obtains the customer ID, generates a monitoring instruction according to the customer ID, and submits the monitoring instruction to the monitoring unit, which then processes the instruction accordingly; The monitoring unit is used to receive monitoring instructions and set a collection cycle, obtain customer behavior within the customer collection cycle, generate a cycle return rate based on the customer behavior, and generate corresponding warning information based on the cycle return rate.

4. The e-commerce operation management system based on big data according to claim 3 is characterized in that: The process of collecting customer behavior includes: Obtain customer behavior type and behavior object, and generate customer behavior; Obtain the collection time of the customer behavior, mark the collection time as the behavior occurrence time, and perform timing to obtain the duration of the behavior; Get the customer ID, and based on the customer ID, send the customer behavior, the time when the behavior occurred, and the duration of the behavior to the database for storage.

5. The e-commerce operation management system based on big data according to claim 4 is characterized in that: The process of obtaining customer behavior within a customer acquisition cycle, generating a cycle return rate according to the customer behavior, and generating corresponding warning information according to the cycle return rate includes: Obtain the quantity corresponding to each behavior type in the customer behavior; The number of customer behaviors whose behavior type is browsing behavior is recorded as browsing value, the number of customer behaviors whose behavior type is purchasing behavior is recorded as purchasing value, and the number of customer behaviors whose behavior type is returning behavior is recorded as returning value; Generate cycle return rate based on browsing value, purchase value and return value; Set return thresholds; If the period return rate is greater than or equal to the return threshold, a first-level warning message is generated based on the customer ID and sent to the warning module, which will take corresponding action; otherwise, no action is required.

6. The e-commerce operation management system based on big data according to claim 5, characterized in that: The process of generating keywords, effective prices, and strategy update cycles for a customer based on their behavior includes: The keywords include strong keywords and weak keywords; The effective price includes an effective price anchor point and an effective price range; Set the behavior update cycle and set the effective time domain according to the behavior update cycle; Acquire customer behaviors within a valid time domain in a database, mark the customer behaviors as valid behaviors, acquire keywords based on the valid behaviors, and send the keywords to a database for storage; Set the duration threshold; Obtain the duration of the behavior corresponding to the valid behavior whose behavior type is browsing behavior. If the duration of the behavior is less than the duration threshold, mark the valid behavior as invalid behavior. Otherwise, no operation is required. Obtaining the behavior type in the valid behavior, and obtaining the behavior object corresponding to the behavior type, and then obtaining the commodity name, commodity major category and commodity minor category corresponding to the behavior object, recording them as vocabulary, and generating a behavior-related vocabulary set; Setting a display threshold; obtaining the percentage of each word according to the behavior-related vocabulary set, and if the percentage is greater than or equal to the display threshold, marking the word as a behavior keyword; if the percentage is less than the display threshold, marking all words as behavior keywords; Mark the behavior keywords corresponding to the behavior types of purchase behavior and browsing behavior as strong keywords, and mark the behavior keywords corresponding to the behavior type of return behavior as weak keywords; If there are overlapping words between strong keywords and weak keywords, delete the overlapping words; According to the effective behavior, obtain the product price corresponding to the strong keyword. If only one product price is obtained, mark the product price as the effective price anchor point. If pr product prices are obtained, generate the effective price range according to the product prices, where pr is greater than 1. Obtain the behavior occurrence time corresponding to the valid behavior whose behavior type is purchase behavior, and mark the occurrence time as the valid purchase time; obtain the behavior occurrence time corresponding to the valid behavior whose behavior type is return behavior, and mark the occurrence time as the valid return time; Establish a one-dimensional timeline and map the effective purchase time and the effective return time to the one-dimensional timeline; Set unit time domain and behavior threshold; If there are a valid purchase times in the time range of the unit time domain mapping on the one-dimensional time axis, then the time range is marked as a valid purchase time zone; where a is greater than the behavior threshold; If there are b valid return times within the time range of the unit time domain mapping on the one-dimensional time axis, then the time range is marked as a valid return time zone; where b is greater than the behavior threshold; Delete the time zones that overlap with the valid return time zones in the valid purchase time zones, and generate valid push time zones; Obtain the start time of each valid push time zone, and generate several time intervals according to the start time; obtain the average value, mode and median corresponding to the several time intervals; Generate a policy update cycle based on the mean, mode, and median.

7. The e-commerce operation management system based on big data according to claim 6, characterized in that: The process of generating internal correlation coefficient and external correlation coefficient based on product information and store information includes: Randomly obtain a commodity and mark it as a target commodity; obtain the internal correlation coefficient and external correlation coefficient between the target commodity and other commodities, wherein the other commodities include internal commodities and external commodities; Obtain a product link corresponding to a target product, and mark the product link as a target link; Obtain a product link in the same product group as the target link, mark the product ID corresponding to the product link as an internal ID; mark the product corresponding to the internal ID as an internal product, and mark the products corresponding to other product IDs in the product unit except the internal ID as external products; Set correlation coefficient_A, correlation coefficient_B and correlation coefficient_C; Obtain the product name, major category, minor category and price of the target product, which are recorded as target product information; obtain the product name, major category, minor category and price of the internal product, which are recorded as internal product information; obtain the product name, major category, minor category and price of the external product, which are recorded as external product information; According to the target product information, the internal product information, the correlation coefficient_A, the correlation coefficient_B and the correlation coefficient_C, the internal correlation coefficient and the external correlation coefficient between the internal product and the target product are obtained; According to the target product information, the internal product information, the external product information, the correlation coefficient_B and the correlation coefficient_C, the internal correlation coefficient and the external correlation coefficient between the external product and the target product are obtained.

8. The e-commerce operation management system based on big data according to claim 7, characterized in that: The process of generating corresponding push information based on keywords, effective prices, internal correlation coefficients and external correlation coefficients includes: Set the positive effective percentage and negative effective percentage, where the positive effective percentage is greater than the negative effective percentage; If the effective price is only one effective price anchor point, the pushed price range = [effective price anchor point × negative effective percentage, effective price anchor point × positive effective percentage]; If the effective price has only one effective price range, then obtain the endpoint values ​​of the effective price range, record them as the left end value and the right end value, and the driving price range = [left end value × negative effective percentage, right end value × positive effective percentage]; Otherwise, according to the principle of generating a push price range from a valid price anchor point and the principle of generating a push price range from a valid price range, a corresponding push price partition is generated through the valid price and the valid percentage, a union between the push price partitions is obtained, and a push price range is generated according to the union; Set the targeting threshold and push amount; Obtain the product name, product category and product subcategory in the product information and record them as information sub-items; if there are num products in the three information sub-items that are consistent with the strong keyword, mark the product as a targeted product; where num is greater than or equal to the targeting threshold; Obtain the internal and external products corresponding to the targeted product; Obtaining an internal correlation coefficient and an external correlation coefficient between the targeted product and the internal product, and generating a push value for the internal product according to the internal correlation coefficient and the external correlation coefficient corresponding to the internal product; Obtaining an internal correlation coefficient and an external correlation coefficient between the target product and the external product, and generating a push value for the external product according to the internal correlation coefficient and the external correlation coefficient corresponding to the external product; The push value = internal correlation coefficient + external correlation coefficient; According to the push value, internal and external products are arranged separately; If there are several commodities that cannot be arranged due to equal push values, they will be arranged according to the commodity prices; if the commodity prices are equal, the commodities will be arranged randomly; After the internal and external products are arranged, the first pushed products whose prices are within the pushed price range are obtained, the products are marked as pushed products, and push information is generated according to the pushed products; wherein push is equal to the push amount value.

9. The e-commerce operation management system based on big data according to claim 8, characterized in that: The process of handling warning information includes: If a first-level warning message is received, query information is generated based on the customer ID, and the query information is submitted to the database to obtain the contact information corresponding to the customer ID. A reminder message is sent to the customer based on the contact information, informing the customer to purchase with caution.

Citation Information

Patent Citations

  • Goods recommendation method and system

    CN103839172A

  • User personalized shopping behavior-based commodity recommendation method

    CN106708821A

  • Internet-based electronic commerce data processing method and system

    CN113888286A

  • Instrument commodity sales early warning method and device, equipment and storage medium

    CN114493667A

  • Financial transaction information management system and method based on big data

    CN116362826A