E-commerce platform management system based on big data

Abstract

The invention relates to the technical field of e-commerce management, and discloses an e-commerce platform management system based on big data, which comprises a data acquisition unit, a preprocessing unit, a data analysis unit and a commodity management unit, and is characterized in that browsing data is carefully processed, invalid data is removed, and deletion and invalid judgment are performed according to comparison between browsing duration and a preset interval; subsequent analysis is ensured to be based on high-quality data, and a foundation is laid for accurately evaluating the attention of the commodity; secondly, commodity management is scientific and efficient; according to the obtained commodity attention degree, comparing with a preset attention degree threshold set, generating different inventory management signals, guiding managers to adjust the inventory level, avoiding overstock or stockout, improving the scientificity and efficiency of inventory management, and improving the operation benefits of an e-commerce platform; scientific inventory management improves supply timeliness and service quality, enhances user experience, improves competitiveness and loyalty, and promotes sustainable development of the platform.

Description

Technical Field

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

[0002] In today's digital age, e-commerce platforms are developing extremely rapidly, with a wide variety of products and a huge number of users. For the management of e-commerce platforms, how to efficiently manage product inventory to meet user needs, improve operational efficiency and reduce costs is crucial.

[0003] Traditional e-commerce platform management systems often rely on experience or simple data analysis in inventory management, and lack accurate analysis of product popularity. It is impossible to accurately understand the degree of attention paid to each product on the platform, and it is difficult to adjust inventory levels according to the actual popularity of the product. For example, for some potentially popular products, insufficient inventory may occur due to the failure to discover their popularity in time, affecting sales; for some products that gradually lose user attention, inventory may be accumulated due to failure to adjust inventory in time, increasing costs.

[0004] In order to solve the above problems, the present invention proposes an e-commerce platform management system based on big data. The system conducts in-depth analysis on the browsing data of each product collected by the e-commerce platform, determines the popularity of the product, and manages the inventory of the e-commerce products based on the popularity, thereby achieving more accurate and efficient e-commerce platform management. Summary of the invention

[0005] The purpose of the present invention is to provide an e-commerce platform management system based on big data, which solves the technical problems raised in the background technology.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] An e-commerce platform management system based on big data, comprising:

[0008] A data collection unit is used to collect browsing data of each product on the e-commerce platform;

[0009] A preprocessing unit, used to preprocess the browsing data and remove invalid data in the browsing data through preprocessing;

[0010] A data analysis unit, used to analyze the pre-processed browsing data and determine the popularity of each product based on the analysis results;

[0011] The product management unit is used to manage the inventory of e-commerce products according to the popularity of each product, and adjust the inventory level of each product based on the inventory management results;

[0012] As a further solution of the present invention: the preprocessing method of the preprocessing unit is as follows:

[0013] SA1. Obtain the browsing time corresponding to each product from the browsing data corresponding to the product by the user in the specified period;

[0014] At the same time, between the start timestamp and the end timestamp corresponding to the corresponding browsing time, obtain the number of clicks on the product interface and the timestamp corresponding to each click;

[0015] SA2. Compare the browsing time with the preset browsing time threshold:

[0016] SA2.1. If the browsing time is greater than or equal to the browsing time threshold, the corresponding browsing times in the browsing data will be deleted;

[0017] SA2.2. If the browsing time is less than the browsing time threshold, the browsing is judged as invalid browsing, and the number of browsing times of the browsing data is reduced by one.

[0018] As a further solution of the present invention: the deletion processing method is as follows:

[0019] SA2.11. Get the number of clicks on the product interface

[0020] Get the number of clicks on the product interface during the browsing time corresponding to the user browsing the product, and mark it as D1;

[0021] SA2.12. Get click interval duration

[0022] At the same time, according to the time trend, the time difference between adjacent click timestamps is obtained and marked as the click interval duration D2;

[0023] SA2.13. Comparison and judgment

[0024] Then, the click interval D1 is compared with the preset time difference threshold D1y, and the number of clicks D2 is compared with the preset number of clicks threshold D2y:

[0025] SA2.14. Determine browsing validity

[0026] When D2≤D2y and D1>D1y, the browsing is judged as invalid browsing, and the browsing number of the browsing data is reduced by one. Otherwise, it means that the browsing is judged as valid browsing, and the browsing number of the browsing data remains unchanged.

[0027] As a further solution of the present invention: the browsing data analysis method of the data analysis unit is as follows:

[0028] 1. Calculate the average browsing interval and standard deviation of products

[0029] SS1. Select a product;

[0030] From the browsing data, obtain the total number of views of the product by all users in a specified period;

[0031] Then, the average browsing interval duration F of the product in the specified period is calculated through F=C / T;

[0032] In the formula, C represents the total number of views of the product in the specified period, and T refers to the time difference in the specified period;

[0033] SS2. At the same time, obtain the timestamp of each browsing from the browsing data within the specified period;

[0034] The timestamp of each browsing is the time node of entering the interface of the relevant product;

[0035] And according to the time trend, get the actual interval between two adjacent browsing timestamps and mark it as J i , i = 1, 2, ... C-1;

[0036] Afterwards, through Calculate the standard deviation P between the actual interval durations within the specified period;

[0037] 2. Marking and encoding the actual interval duration and calculating the proportion of the number

[0038] SS3, based on F and P to J i Add tag code;

[0039] The marking code is preset, and the marking code includes "0", "1", "2" and "3";

[0040] Select a J i ;

[0041] If | FJ i |≤P, then the J i Add a predetermined marker code "0";

[0042] If P<|FJ i |≤2P, then the J i Add a predetermined mark code "1";

[0043] If 2P<|FJ i |≤3P, then the J i Add the predetermined tag code "2";

[0044] If | FJ i |>3P, then the J i Add the predetermined mark code "3";

[0045] SS4, in all J i In the above equation, extract the number u0 of Ji corresponding to the mark code “0”;

[0046] And by B0=u0 / (C-1), calculate the proportion B0 of the number of Ji corresponding to the mark code "0" in the specified period;

[0047] At the same time, the number of Ji u1, u2 and u3 corresponding to the tag codes "1", "2" and "3" are extracted in sequence;

[0048] And according to the calculation method of B0, it is obtained that the number of Ji corresponding to the mark code "1", "2" and "3" in the specified period accounts for B1, B2 and B3 respectively;

[0049] Then, among B0, B1, B2 and B3, the percentage of the largest value is obtained;

[0050] If the largest value accounts for B0, a predetermined mark score β1 is added to the product;

[0051] If the largest value accounts for B1, a predetermined mark score β2 is added to the product;

[0052] If the largest value accounts for B2, a predetermined mark score β3 is added to the product;

[0053] If the largest value accounts for B3, a predetermined mark score β4 is added to the product;

[0054] Among them, β1>β2>β3>β4, and the value of the marking score is the preset value;

[0055] 3. Mark and encode browsing time and calculate the proportion of browsing time

[0056] SS5. At the same time, within the specified period, obtain the browsing duration of each browsing from the browsing data and mark it as L j , j = 1, 2, ... C;

[0057] Afterwards through Calculate the average browsing time LP of the product in the specified period;

[0058] Then through Calculate the standard deviation Lb between the browsing durations of each browsing within a specified period;

[0059] SS6, add marking code to Lj according to Lp and Lb;

[0060] The marking code is added in the same way as that of SS3, and the marking codes are "0", "1", "2" and "3".

[0061] The method is: select an L j ;

[0062] If | FL j |≤P, then the L j Add a predetermined marker code "0";

[0063] If P<|FL j |≤2P, then the L j Add a predetermined mark code "1";

[0064] If 2P<|FL j |≤3P, then the L j Add the predetermined tag code "2";

[0065] If | FL j |>3P, then the L j Add the predetermined mark code "3";

[0066] SS7, at the same time, according to the SS4 method, extract the mark codes "0", "1", "2" and "3" in turn, corresponding to L j The number v0, v1, v2 and v3, and the corresponding L j The number of percentages are marked as H0, H1, H2 and H3 respectively;

[0067] Then, in H0, H1, H2 and H3, the largest value is obtained;

[0068] If the largest value accounts for H0, then a predetermined mark score β1 is added to the product;

[0069] If the largest value accounts for H1, a predetermined mark score β2 is added to the product;

[0070] If the largest value is H2, a predetermined mark score β3 is added to the product;

[0071] If the largest value is H3, a predetermined mark score β4 is added to the product;

[0072] Among them, the values ​​corresponding to β1, β2, β3, and β4 in this step are consistent with the values ​​of SS4;

[0073] 4. Calculate the popularity of products

[0074] SS8, obtain the labeling scores of SS4 and SS7, and label them as βa and βb respectively;

[0075] Then, the popularity of the product K is calculated through K = βa × α1 + βb × α2;

[0076] Among them, α1 and α2 are preset weight factors;

[0077] SS9. According to the method of SS1 to SS8, calculate the popularity of all commodities and mark it as Kg, where g = 1, 2, ... e, and e represents the quantity of all commodities.

[0078] As a further solution of the present invention: the inventory management method of the commodity management unit is as follows:

[0079] Compare the attention Kg of each product with the preset attention threshold set YK∈[YK1,YK2]:

[0080] If Kg < YK1, it means that the popularity of the corresponding product is low, and an inventory management signal 1 is generated. The inventory management signal 1 is used to remind relevant managers to consider reducing the inventory level;

[0081] If YK1≤Kg<YK2, it means that the popularity of the corresponding product is average, and the second inventory management signal is generated. The second inventory management signal is used to remind the relevant management personnel to consider maintaining a moderate inventory level for the corresponding product;

[0082] If Kg≥YK2, it means that the popularity of the corresponding product is high, and an inventory management signal 1 is generated. The inventory management signal 1 is used to remind relevant managers to consider increasing the inventory level.

[0083] Beneficial effects of the present invention:

[0084] In the present invention, the data analysis unit adopts a multi-dimensional analysis method, comprehensively considering the average browsing interval duration, standard deviation, and different tag codes and scores, and can more comprehensively and accurately evaluate the popularity of the product. For example, through the detailed analysis of the browsing timestamp and browsing duration, an accurate data basis is provided for the calculation of the popularity of the product.

[0085] In the present invention, the commodity management unit performs inventory management based on the popularity of the commodity, so that the inventory level can be matched with the actual demand for the commodity. By comparing the popularity with a preset attention threshold set, different inventory management signals are generated to remind the management personnel to adopt corresponding inventory adjustment strategies, avoid inventory backlogs or out-of-stock situations, and improve the scientificity and efficiency of inventory management.

[0086] The present invention can timely remind customers to reduce inventory levels for low-popularity products, thereby reducing inventory costs and risks; it can maintain a moderate inventory level for products with average popularity, thereby meeting certain market demands without causing excessive inventory backlogs; and it can consider increasing inventory levels for products with high popularity, thereby ensuring that market demands can be met, thereby increasing sales opportunities and customer satisfaction.

[0087] The present invention, based on the analysis and management system of big data, can help e-commerce platforms better understand the market performance and user needs of commodities, and provide strong support for the operation decision-making of the platform. By accurately evaluating the popularity of commodities and conducting reasonable inventory management, the efficiency of capital use can be improved, the operating costs can be reduced, and the overall economic benefits of the platform can be improved.

[0088] The present invention processes browsing data through a preprocessing unit, removes invalid data, and ensures that subsequent analysis is based on valid information. Specifically, by comparing the browsing time with a preset interval, deleting the number of browsing times and determining invalid browsing, valuable browsing data can be accurately screened out, thereby improving the quality and reliability of the data.

[0089] The present invention, scientific inventory management can also improve the timeliness of commodity supply and service quality, enhance user experience, improve the competitiveness and user loyalty of the platform, and promote the sustainable development of the e-commerce platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] The present invention will be further described below in conjunction with the accompanying drawings.

[0091] Figure 1 It is a system block diagram of an e-commerce platform management system based on big data in the present invention.

[0092] Figure 2 It is a unit flow diagram of a data analysis unit in an e-commerce platform management system based on big data of the present invention.

[0093] Figure 3 It is a flow chart of deletion processing in a data analysis unit in an e-commerce platform management system based on big data according to the present invention. DETAILED DESCRIPTION

[0094] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0095] Embodiment 1

[0096] See also Figure 1 and Figure 3 As shown, the present invention is an e-commerce platform management system based on big data, comprising:

[0097] A data collection unit is used to collect browsing data of each product on the e-commerce platform;

[0098] A data analysis unit, used for performing browsing data analysis on the browsing data;

[0099] The browsing data analysis method is as follows:

[0100] SS1, take a product as an example;

[0101] From the browsing data, obtain the total number of views of the product by all users in a specified period;

[0102] Then, the average browsing interval duration F of the product in the specified period is calculated through F=C / T;

[0103] In the formula, C represents the total number of views of the product in the specified period, and T refers to the time difference in the specified period;

[0104] SS2. At the same time, obtain the timestamp of each browsing from the browsing data within the specified period;

[0105] The timestamp of each browsing is the time node of entering the interface of the relevant product;

[0106] And according to the time trend, get the actual interval between two adjacent browsing timestamps and mark it as J i , i = 1, 2, ... C-1;

[0107] Afterwards, through Calculate the standard deviation P between the actual interval durations within the specified period;

[0108] SS3, based on F and P to J i Add tag code;

[0109] The marking code is preset, and the marking code includes "0", "1", "2" and "3";

[0110] Select a J i ;

[0111] If | FJ i |≤P, then the J i Add a predetermined marker code "0";

[0112] If P<|FJ i |≤2P, then the J iAdd a predetermined mark code "1";

[0113] If 2P<|FJ i |≤3P, then the J i Add the predetermined tag code "2";

[0114] If | FJ i |>3P, then the J i Add the predetermined mark code "3";

[0115] SS4, in all J i In the above equation, extract the number u0 of Ji corresponding to the mark code “0”;

[0116] And by B0=u0 / (C-1), calculate the proportion B0 of the number of Ji corresponding to the mark code "0" in the specified period;

[0117] At the same time, the number of Ji u1, u2 and u3 corresponding to the tag codes "1", "2" and "3" are extracted in sequence;

[0118] And according to the calculation method of B0, it is obtained that the number of Ji corresponding to the mark code "1", "2" and "3" in the specified period accounts for B1, B2 and B3 respectively;

[0119] Then, among B0, B1, B2 and B3, the percentage of the largest value is obtained;

[0120] If the largest value accounts for B0, a predetermined mark score β1 is added to the product;

[0121] If the largest value accounts for B1, a predetermined mark score β2 is added to the product;

[0122] If the largest value accounts for B2, a predetermined mark score β3 is added to the product;

[0123] If the largest value accounts for B3, a predetermined mark score β4 is added to the product;

[0124] Wherein, β1>β2>β3>β4, the value of the marking score is a preset value. In this embodiment, β1 is 90, β2 is 70, β3 is 50, and β4 is 30;

[0125] The commodity management unit is used to manage the inventory of e-commerce commodities according to the mark scores of each commodity;

[0126] The product's label score is labeled as βk;

[0127] Compare the tag score βk of each product with the preset attention threshold set Yt∈[Yt1,Yt2]:

[0128] If βk<Yt1, it means that the popularity of the corresponding product is low, and an inventory management signal 1 is generated. The inventory management signal 1 is used to remind relevant managers to consider reducing the inventory level;

[0129] If Yt1≤βk<Yt2, it means that the popularity of the corresponding product is average, and the second inventory management signal is generated. The second inventory management signal is used to remind the relevant management personnel to consider maintaining a moderate inventory level for the corresponding product;

[0130] If βk≥Yt2, it means that the popularity of the corresponding product is high, and an inventory management signal 1 is generated. The inventory management signal 1 is used to remind relevant managers to consider increasing inventory levels.

[0131] This embodiment analyzes the product browsing data, calculates the average browsing interval duration and standard deviation, marks and encodes the browsing timestamps, and obtains different marking scores, thereby providing a quantitative basis for product management. The inventory of e-commerce products is managed based on the marking scores of the products, which can remind management personnel to adjust the inventory level according to the popularity of the products, improve the scientificity and rationality of inventory management, avoid inventory backlogs or out-of-stock situations, reduce costs, and improve operational efficiency.

[0132] Embodiment 2

[0133] As the second embodiment of the present invention, when the present application is implemented, compared with the first embodiment, the technical solution of the present embodiment is different from that of the first embodiment only in that,

[0134] The browsing data analysis method further includes the following steps based on the first embodiment:

[0135] SS5. At the same time, within the specified period, obtain the browsing duration of each browsing from the browsing data and mark it as L j , j = 1, 2, ... C;

[0136] Afterwards through Calculate the average browsing time LP of the product in the specified period;

[0137] Then through Calculate the standard deviation Lb between the browsing durations of each browsing within a specified period;

[0138] SS6, add marking code to Lj according to Lp and Lb;

[0139] The marking code is added in the same way as that of SS3, and the marking codes are "0", "1", "2" and "3".

[0140] The method is: select an L j ;

[0141] If | FL j |≤P, then the L j Add a predetermined marker code "0";

[0142] If P<|FL j |≤2P, then the L j Add a predetermined mark code "1";

[0143] If 2P<|FL j |≤3P, then the L j Add the predetermined tag code "2";

[0144] If | FL j |>3P, then the L j Add the predetermined mark code "3";

[0145] SS7. At the same time, according to the method of SS4 in Example 1, the mark codes "0", "1", "2" and "3" are extracted in sequence, corresponding to L j The number v0, v1, v2 and v3, and the corresponding L j The number of percentages are marked as H0, H1, H2 and H3 respectively;

[0146] Then, in H0, H1, H2 and H3, the largest value is obtained;

[0147] If the largest value accounts for H0, then a predetermined mark score β1 is added to the product;

[0148] If the largest value accounts for H1, a predetermined mark score β2 is added to the product;

[0149] If the largest value is H2, a predetermined mark score β3 is added to the product;

[0150] If the largest value is H3, a predetermined mark score β4 is added to the product;

[0151] In this embodiment, the values ​​corresponding to β1, β2, β3, and β4 in this step are consistent with the value of SS4;

[0152] SS8, obtain the labeling scores of SS4 and SS7, and label them as βa and βb respectively;

[0153] Then, the popularity of the product K is calculated through K = βa × α1 + βb × α2;

[0154] Among them, α1 and α2 are preset weight factors;

[0155] SS9. Calculate the popularity of all products in the same way as SS1 to SS8 and mark them as Kg, where g = 1, 2, ... e, and e represents the number of all products.

[0156] The product management unit is used to manage the inventory of e-commerce products based on the popularity of each product;

[0157] Compare the attention Kg of each product with the preset attention threshold set YK∈[YK1,YK2]:

[0158] If Kg < YK1, it means that the popularity of the corresponding product is low, and an inventory management signal 1 is generated. The inventory management signal 1 is used to remind relevant managers to consider reducing the inventory level;

[0159] If YK1≤Kg<YK2, it means that the popularity of the corresponding product is average, and the second inventory management signal is generated. The second inventory management signal is used to remind the relevant management personnel to consider maintaining a moderate inventory level for the corresponding product;

[0160] If Kg≥YK2, it means that the popularity of the corresponding product is high, and an inventory management signal 1 is generated. The inventory management signal 1 is used to remind relevant managers to consider increasing the inventory level.

[0161] Based on the first embodiment, this embodiment adds analysis of the browsing time of each time, calculates the average browsing time and standard deviation, and performs marking coding and scoring, so as to make the evaluation of the popularity of the product more comprehensive; by comprehensively considering the two factors of browsing interval and browsing time to calculate the popularity of the product, a more accurate basis is provided for inventory management, the inventory management strategy is further optimized, and the degree of refinement of inventory management is improved.

[0162] Embodiment 3

[0163] As the third embodiment of the present invention, when the present application is implemented, compared with the first and second embodiments, the technical solution of this embodiment is to combine the solutions of the first and second embodiments, and the difference between the technical solution of this embodiment and the first and second embodiments is that in this embodiment,

[0164] The commodity management unit is also used to manage the inventory of e-commerce commodities based on the tag scores of each commodity obtained by SS7;

[0165] The label score of each product obtained by SS7 is βq;

[0166] Compare the tag score βq of each product with the preset attention threshold set Yt∈[Yt1,Yt2]:

[0167] If βq<Yt1, it means that the popularity of the corresponding product is low, and an inventory management signal 1 is generated. The inventory management signal 1 is used to remind relevant managers to consider reducing the inventory level;

[0168] If Yt1≤βq<Yt2, it means that the popularity of the corresponding product is average, and the second inventory management signal is generated. The second inventory management signal is used to remind relevant managers to consider maintaining a moderate inventory level for the corresponding product;

[0169] If βq≥Yt2, it means that the popularity of the corresponding product is high, and an inventory management signal 1 is generated. The inventory management signal 1 is used to remind relevant managers to consider increasing inventory levels.

[0170] This embodiment combines the solutions of embodiment one and embodiment two to conduct a more comprehensive evaluation of the goods, and at the same time manages the inventory of e-commerce goods based on the tag scores in the combined solution, thereby improving the accuracy and reliability of inventory management decisions; it can more accurately adjust the inventory level according to the actual situation of the goods, meet market demand, and improve the economic benefits of the enterprise.

[0171] Embodiment 4

[0172] As the fourth embodiment of the present invention, when the present application is implemented, compared with the first embodiment, the second embodiment and the third embodiment, the difference between the present embodiment and the first embodiment, the second embodiment and the third embodiment is that the present embodiment further includes:

[0173] A preprocessing unit, used for preprocessing browsing data;

[0174] The preprocessing method is as follows:

[0175] SA1. Obtain the browsing time corresponding to each product from the browsing data corresponding to the product by the user within the specified period;

[0176] At the same time, between the start timestamp and the end timestamp corresponding to the corresponding browsing time, obtain the number of clicks on the product interface and the timestamp corresponding to each click;

[0177] SA1. Compare the browsing time with the preset browsing time threshold:

[0178] If the browsing time is less than the browsing time threshold, the browsing is determined to be invalid, and the number of browsing data is reduced by one;

[0179] If the browsing time is greater than or equal to the browsing time threshold, the number of clicks on the product interface during the browsing time corresponding to the user browsing the product is obtained and marked as D1;

[0180] At the same time, according to the time trend, the time difference between adjacent click timestamps is obtained and marked as the click interval duration D2;

[0181] Then, the click interval D1 is compared with the preset time difference threshold D1y, and the number of clicks D2 is compared with the preset number of clicks threshold D2y:

[0182] When D2≤D2y and D1>D1y, the browsing is judged as invalid browsing, and the browsing number of the browsing data is reduced by one. Otherwise, it means that the browsing is judged as valid browsing, and the browsing number of the browsing data remains unchanged.

[0183] This embodiment adds a preprocessing unit to preprocess the browsing data, removes invalid browsing data, and improves the accuracy and reliability of the data; by judging the browsing time, number of clicks, and click interval length, it ensures that the browsing data used for analysis is more real and effective, providing a more reliable basis for subsequent data analysis and inventory management.

[0184] Embodiment 5

[0185] As the fifth embodiment of the present invention, when the present application is specifically implemented, compared with the first, second, third and fourth embodiments, the technical solution of this embodiment is to combine and implement the solutions of the above-mentioned first, second, third and fourth embodiments.

[0186] This embodiment combines the solutions of Embodiment 1, Embodiment 2, Embodiment 3 and Embodiment 4 to achieve comprehensive analysis and preprocessing of product browsing data, improve data quality and the accuracy of analysis results; and provide a complete, scientific and effective solution for inventory management of e-commerce products, which can better meet the inventory management needs of enterprises and improve the competitiveness and operational efficiency of enterprises.

[0187] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0188] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An e-commerce platform management system based on big data, characterized in that: include: A data analysis unit is used to analyze the browsing data of each product collected from the e-commerce platform, and determine the popularity of each product based on the analysis results; The product management unit is used to manage the inventory of e-commerce products based on the popularity of each product, and adjust the inventory level of each product based on the inventory management results.

2. According to the big data-based e-commerce platform management system of claim 1, it is characterized in that: The browsing data analysis method of the data analysis unit is as follows: SS1. Select a product and obtain the total number of views of the product by all users within a specified period from the browsing data. Then, the average browsing interval of the product in the specified period is calculated by dividing the total number of browsing times by the time difference in the specified period, and marked as F; SS2. At the same time, obtain the timestamp of each browsing from the browsing data within the specified period; The timestamp of each browsing is the time node of entering the interface of the relevant product; And according to the time trend, get the actual interval between two adjacent browsing timestamps and mark it as J i , i = 1, 2, ... C-1; where C represents the total number of views of the product in the specified period; After that, the standard deviation between the actual interval durations within the specified period is calculated using the standard deviation formula and marked as P; SS3, based on F and P to J i Add tag code; SS4, in all J i , extract the number of Ji corresponding to each tag code; Then calculate the proportion of the number of Ji corresponding to each tag code in C-1 within the specified period; Afterwards, the quantity proportion with the largest value is obtained from the quantity proportions corresponding to each mark code, and a corresponding predetermined mark score is added to the product; SS5. At the same time, within the specified period, obtain the browsing duration of each browsing from the browsing data and mark it as L j , j = 1, 2, ... C; After that, the average browsing time of the product in the specified period is calculated and marked as LP; Next, the standard deviation of the browsing duration of each browsing in the specified period is calculated using the standard deviation formula and marked as Lb; SS6, based on SS5, adds Lj with marking code according to Lp and Lb; The marking code is added in the same way as that of SS3. SS7: Based on SS6 and in the same manner as SS4, add a corresponding predetermined mark score to the product; SS8, obtain the labeling scores of SS4 and SS7, and label them as βa and βb respectively; Then, the popularity of the product K is calculated through K = βa × α1 + βb × α2; Among them, α1 and α2 are preset weight factors; SS9. According to the method of SS1 to SS8, calculate the popularity of all commodities and mark it as Kg, where g = 1, 2, ... e, and e represents the quantity of all commodities.

3. According to the big data-based e-commerce platform management system of claim 2, it is characterized in that: In SS3, the tag codes include "0", "1", "2" and "3", and they are pre-set code values; The tag encoding in SS3 is added as follows: Select a J i ; If | FJ i |≤P, then the J i Add the predetermined marker code "0"; If P<|FJ i |≤2P, then the J i Add a predetermined marker code "1"; If 2P<|FJ i |≤3P, then the J i Add the predetermined tag code "2"; If | FJ i |>3P, then the J i Add the predetermined tag code "3".

4. According to claim 3, an e-commerce platform management system based on big data is characterized in that: In SS4, the number of quantity proportions is determined based on the number of marker codes; According to the marking codes "0", "1", "2" and "3", the corresponding quantity proportions are marked as B0, B1, B2, B3; If the largest value accounts for B0, a predetermined mark score β1 is added to the product; If the largest value accounts for B1, a predetermined mark score β2 is added to the product; If the largest value accounts for B2, a predetermined mark score β3 is added to the product; If the largest value accounts for B3, a predetermined mark score β4 is added to the product; Among them, β1>β2>β3>β4, and the value of the marking score is a preset value.

5. The e-commerce platform management system based on big data according to claim 4 is characterized in that: in, The corresponding values ​​of β1, β2, β3, and β4 in SS7 are consistent with the values ​​of β1, β2, β3, and β4 in SS4.

6. The e-commerce platform management system based on big data according to claim 2, characterized in that: The inventory management method of the commodity management unit is as follows: Compare the attention Kg of each product with the preset attention threshold set YK∈[YK1,YK2]: If Kg < YK1, it means that the popularity of the corresponding product is low, and an inventory management signal 1 is generated. The inventory management signal 1 is used to remind relevant managers to consider reducing the inventory level; If YK1≤Kg<YK2, it means that the popularity of the corresponding product is average, and the second inventory management signal is generated. The second inventory management signal is used to remind the relevant management personnel to consider maintaining a moderate inventory level for the corresponding product; If Kg≥YK2, it means that the popularity of the corresponding product is high, and an inventory management signal 1 is generated. The inventory management signal 1 is used to remind relevant managers to consider increasing the inventory level.

7. The e-commerce platform management system based on big data according to claim 1, characterized in that: Also includes: The data collection unit is used to collect browsing data of each product on the e-commerce platform.

8. The e-commerce platform management system based on big data according to claim 1, characterized in that: Also includes: A preprocessing unit, used to preprocess the browsing data of each product collected from the e-commerce platform, and remove invalid data in the browsing data through preprocessing; The data analysis unit is used to perform browsing data analysis on the pre-processed browsing data.

9. The e-commerce platform management system based on big data according to claim 8, characterized in that: The preprocessing method of the preprocessing unit is as follows: SA1. Obtain the browsing time corresponding to each product from the browsing data corresponding to the product by the user within the specified period; At the same time, between the start timestamp and the end timestamp corresponding to the corresponding browsing time, obtain the number of clicks on the product interface and the timestamp corresponding to each click; SA2. Compare the browsing time with the preset browsing time threshold: SA2.

1. If the browsing time is greater than or equal to the browsing time threshold, the corresponding browsing times in the browsing data will be deleted; SA2.

2. If the browsing time is less than the browsing time threshold, the browsing is judged as invalid browsing, and the number of browsing times of the browsing data is reduced by one.

10. The e-commerce platform management system based on big data according to claim 9, characterized in that: The deletion process is as follows: SA2.

11. Obtain the number of clicks on the product interface during the browsing time corresponding to the user browsing the product, and mark it as D1; SA2.

12. At the same time, according to the time trend, obtain the time difference between adjacent click timestamps and mark it as the click interval duration D2; SA2.13, then compare the click interval D1 with the preset time difference threshold D1y, and the click count D2 with the preset click count threshold D2y: SA2.

14. When D2≤D2y and D1>D1y, the browsing is judged as invalid browsing, and the number of browsing times of the browsing data is reduced by one. Otherwise, it means that the browsing is judged as valid browsing, and the number of browsing times of the browsing data remains unchanged.

Images