A marketing method for online sales

Abstract

The present invention relates to the technical field of online marketing, and discloses a marketing method for online sales, comprising: obtaining user's favorite products, dividing a day into multiple time periods, and calculating the user's historical average browsing time in each time period; judging in real time whether the user has opened a shopping software, and if the user has opened it, obtaining the current time period and calculating the total number and total time of the user's browsed products to determine the browsing time of a single product; recommending a specific number of products from a set of intended products according to the user's browsing time and average browsing time of a single product in the current time period; obtaining the user's browsing status of the recommended products in real time, and adjusting subsequent recommended products according to the user's browsing feedback.

Description

Technical Field

[0001] The present invention relates to the technical field of online marketing, and in particular to a marketing method for online sales. Background Art

[0002] With the rapid development of e-commerce, online sales competition has become increasingly fierce. Merchants need to continuously innovate and optimize their marketing strategies to improve user purchase conversion rates and customer satisfaction. In this context, personalized recommendation methods based on user behavior analysis have gradually become one of the important tools for online sales.

[0003] Traditional product recommendation methods often rely on static user data and simple rules, which are difficult to adapt to the dynamic changes in user behavior. Users' browsing habits and preferences may vary significantly in different time periods. Therefore, how to obtain and analyze users' browsing data in real time and provide accurate product recommendations has become a hot topic of research.

[0004] The present invention proposes an online sales and marketing method based on user browsing behavior, which executes marketing strategies by acquiring user-favorite products, dividing time periods and analyzing historical data, real-time data monitoring and recommendation, and integrating and adjusting user feedback. This not only improves the personalization of product recommendations, but also enhances the user's shopping experience. This data-driven marketing strategy meets the needs of modern consumers for personalized services and has broad application prospects. Summary of the invention

[0005] The present invention provides a marketing method for online sales, which is used to promote the solution of the problems mentioned in the above background technology.

[0006] The present invention provides the following technical solution: a marketing method for online sales, comprising:

[0007] S1. Obtain the user's favorite products and enter the favorite products into the intended product set;

[0008] S2. Divide the time period, obtain the historical data of the user in each time period, and calculate the average browsing time corresponding to each time period based on the historical data of each time period;

[0009] The historical data specifically refers to the usage time from the start time to the end time each time the user uses the shopping software;

[0010] S3. If the user opens the shopping software, the time period is determined according to the opening time, the total number and total time of the products browsed by the user in the current time period are obtained, and the browsing time of a single product by the user is calculated;

[0011] Recommend products based on the user's browsing time and average browsing time for a single product in the current time period;

[0012] S4, selecting products according to the number of recommended products n and making the first recommendation, obtaining the actual number of products browsed by the user during the current actual usage time, and determining whether to enter the unbrowsed products in the recommended products into the pre-selected set;

[0013] S5. Update the user's relevant data based on the user's browsing of the recommended products;

[0014] Relevant data include users’ total browsing time and average browsing time;

[0015] S6. When the user opens the shopping software again, product recommendations are made based on the pre-selected set and the number m of products selected when the user opens the shopping software again;

[0016] Again refers to the number of times users open the shopping app after the first recommendation.

[0017] Optionally, obtaining the user's favorite product and entering the favorite product into the intended product set includes:

[0018] Get all the items in the user's shopping cart and enter the desired item set;

[0019] Set the quantity threshold a;

[0020] If the number of elements in the intended product set is less than a, the user's historical search keywords are obtained;

[0021] Get all products that match the historical search keywords and enter the intended product set.

[0022] Optionally, dividing the time periods, obtaining historical data of the user in each time period, and calculating the average browsing time corresponding to each time period according to the historical data of each time period, includes:

[0023] Divide 24 hours into d time periods, and the length of each time period is equal;

[0024] Obtain the duration of each time the user uses the shopping app in history, and record it as browsing time;

[0025] Record browsing time into user browsing collection;

[0026] Divide all browsing times in the user browsing set into multiple sub-user browsing sets according to the time period at the start time;

[0027] The start time is the start time of the browsing time;

[0028] Randomly select the sub-user browsing set as the target set;

[0029] Calculate the sum of all browsing time in the target set and record it as the total browsing time of the current target set;

[0030] Calculate the ratio between the total browsing time and the number of elements in the target collection;

[0031] The ratio is recorded as the average browsing time of the time period to which the target set belongs;

[0032] Reselect the sub-user browsing set as the new target set, and calculate the average browsing time of the time period to which the target set belongs;

[0033] When all sub-user browsing sets are traversed, the selection calculation stops.

[0034] Optionally, obtaining the total number of products and the total time that the user browses in the current time period, calculating the user's browsing time for a single product, and determining the number of recommended products n according to the browsing time for a single product and the average browsing time in the user's current time period includes:

[0035] If the user opens the shopping app, the time when the user opens the shopping app is obtained and recorded as the start time;

[0036] Determine the time period in which the start time falls, and obtain the sub-user browsing set corresponding to the time period;

[0037] Record the sub-user browsing set corresponding to the time period as the user monitoring set;

[0038] Get the total browsing time of the history corresponding to the user monitoring collection;

[0039] Calculate the sum of the number of browsed products at each browsing time in the user monitoring set, and record it as the total number of browsed products;

[0040] Calculate the ratio between the total browsing time of the user monitoring set and the total number of browsings, and record it as the browsing time of a single product;

[0041] The ratio of the average browsing time of the user monitoring set to the browsing time of a single product is recorded as the number of product recommendations n.

[0042] Optionally, selecting products according to the number of recommended products n and making the first recommendation, obtaining the actual number of products browsed by the user in the current actual usage time, and determining whether to enter the unbrowsed products in the recommended products into the pre-selected set include:

[0043] Randomly select a product from the set of n intended products as the predicted product;

[0044] Recommend the selected products to the user and remove the selected products from the intended product set;

[0045] Get the time interval of the user's current use of the shopping software and record it as the actual use time;

[0046] Get the number of products browsed by the user during the actual usage time, which is recorded as the actual number of products;

[0047] If the actual number of products is less than n, the unbrowsed products in the predicted products are marked and entered into the pre-selected set;

[0048] If the actual number of products is greater than or equal to n, no entry operation is performed on the pre-selected set.

[0049] Optionally, updating the user's relevant data according to the user's browsing of the recommended products includes:

[0050] The actual usage time is recorded in the sub-user browsing set of the time period;

[0051] Recalculate the sum of the elements in the current sub-user's browsing collection and use it as the total browsing time of the current sub-user's browsing collection;

[0052] Calculate the ratio between the total browsing time and the number of elements in the current sub-user's browsing set;

[0053] The ratio is recorded as the average browsing time of the current sub-user browsing collection in the time period.

[0054] Optionally, when the user opens the shopping software again, the product recommendation is performed according to the pre-selected set and the number m of products selected corresponding to when the user opens the shopping software again, including:

[0055] If the user opens the shopping software again, the time when the user opens the shopping software is obtained and recorded as the first time;

[0056] Determine the time period of the first time, and obtain the sub-user browsing set corresponding to the time period;

[0057] Record the sub-user browsing set corresponding to the time period as the target set;

[0058] Get the total browsing time of the target collection;

[0059] Calculate the sum of the number of browsed products at each browsing time in the target set, and record it as the first value;

[0060] Calculate the ratio between the total browsing time of the target collection and the first value, and record it as the second value;

[0061] The ratio of the average browsing time of the target set to the second value is recorded as the number of selected products m;

[0062] If the number of elements in the pre-selected set is greater than or equal to m, the user's purchase amount each time is obtained and entered into the consumption set;

[0063] Get the maximum and minimum values ​​in the consumption set and record them as the maximum amount and minimum amount;

[0064] A user consumption capacity interval is formed according to the maximum amount and the minimum amount, specifically, the minimum amount is the minimum value of the user consumption capacity interval, and the maximum amount is the maximum value of the user consumption capacity interval;

[0065] Enter all elements in the pre-selected set whose prices are within the user's spending capacity into the preferred set;

[0066] If the number of preferred sets is greater than or equal to m, randomly select m items from the preferred sets and enter them into the recommended item set, and recommend all items in the recommended item set to the user;

[0067] Eliminate the elements in the pre-selected set that are the same as the product recommendation set;

[0068] If the number of preferred sets is less than m, all elements in the preferred set are entered into the product recommendation set;

[0069] Calculate the difference between the number of optimal sets and m, and record it as the remaining number k;

[0070] Randomly select k items that are not selected in the pre-selected set and enter them into the product recommendation set;

[0071] Recommend all elements in the product recommendation set to the user;

[0072] Eliminate the elements in the pre-selected set that are the same as the product recommendation set.

[0073] Optionally, when the user opens the shopping software again, the product recommendation is performed according to the pre-selected set and the number m of products selected corresponding to when the user opens the shopping software again, including:

[0074] If the number of elements in the pre-selected set is less than m, all elements in the pre-selected set are entered into the product recommendation set;

[0075] Calculate the difference between m and the number of elements in the pre-selected set, and record the difference as c;

[0076] Enter all elements in the intended product set whose prices are within the user's consumption capacity range into the qualified product range;

[0077] Randomly select c elements from the qualified product set and enter them into the product recommendation set;

[0078] Recommend all elements in the product recommendation set to the user;

[0079] Eliminate the elements in the pre-selected set and the intended product set that are the same as the product recommendation set.

[0080] The present invention has the following beneficial effects:

[0081] 1. This is a marketing method used for online sales. By recording users' favorite products and building a collection of intended products, it can ensure that the recommended products meet users' preferences and increase users' click-through rate and purchase rate.

[0082] 2. This marketing method used for online sales uses historical data from different time periods to analyze users' browsing habits and the average browsing time of individual products, which helps to make personalized recommendations based on the user's time period characteristics and increase the effectiveness of the recommendations.

[0083] 3. This is a marketing method for online sales. It determines in real time the time period when the user opens the shopping software, obtains the corresponding user historical browsing information according to the time period, calculates the user's browsing time of a single product in the current time period through the historical browsing information, and recommends an appropriate number of products to the user based on the average browsing time and the browsing time of a single product in the current time period. By obtaining the browsing situation of the user in the current time period and making recommendations based on the current browsing time and the average browsing time of the product, it can better adapt to the user's real-time needs and improve the user experience and conversion rate.

[0084] 4. This is a marketing method used for online sales. By updating and optimizing the data of user browsing behavior and adjusting the recommendation model according to the latest browsing data each time, the accuracy of product recommendations can be continuously improved.

[0085] 5. This marketing method used for online sales combines multi-dimensional information such as user browsing behavior, consumption ability, product prices, etc. to ensure that the recommended products not only meet the user's browsing habits, but also match the user's consumption ability range, thereby increasing the possibility of conversion.

[0086] 6. This is a marketing method for online sales. By obtaining the user's browsing situation for recommended products, the unbrowsed products are entered into a pre-selected set for priority recommendation next time. When the number of elements in the pre-selected set is insufficient for the next recommendation, the user's consumption capacity range is obtained through the user's historical consumption record. Based on the user's consumption capacity range, the products that meet the user's consumption capacity are recommended first. By adjusting the browsing situation of the recommended products, the recommended product set is continuously optimized to ensure that the number of products in the recommended product set is sufficient and meets the user's consumption capacity, thereby improving the accuracy of recommendations and the efficiency of product display.

[0087] 7. This marketing method for online sales can capture the specific behavioral characteristics of users by dividing their browsing behavior into multiple time periods and using different recommendation strategies for different time periods, thereby optimizing the recommendation effect and improving the relevance of recommended products and user satisfaction.

[0088] 8. This is a marketing method used for online sales. By randomly selecting products to predict user behavior and adjusting the recommendation strategy, it helps to promptly correct recommended products that do not hit the user's interests, thereby improving the hit rate of product recommendations and user retention rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0090] 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.

[0091] Example 1, see Figure 1 , a marketing method for online sales, including:

[0092] S1. Obtain the user's favorite products and enter the favorite products into the intended product set;

[0093] S2. Divide the time period, obtain the historical data of the user in each time period, and calculate the average browsing time corresponding to each time period based on the historical data of each time period;

[0094] The historical data specifically refers to the usage time from the start time to the end time each time the user uses the shopping software;

[0095] S3. If the user opens the shopping software, the time period is determined according to the opening time, the total number and total time of the products browsed by the user in the current time period are obtained, and the browsing time of a single product by the user is calculated;

[0096] Determine the number of recommended products n based on the browsing time and average browsing time of a single product in the user's current time period;

[0097] S4, selecting products according to the number of recommended products n and making the first recommendation, obtaining the actual number of products browsed by the user during the current actual usage time, and determining whether to enter the unbrowsed products in the recommended products into the pre-selected set;

[0098] S5. Update the user's relevant data based on the user's browsing of the recommended products;

[0099] Relevant data include users’ total browsing time and average browsing time;

[0100] S6. When the user opens the shopping software again, product recommendations are made based on the pre-selected set and the number m of products selected when the user opens the shopping software again;

[0101] Again refers to the number of times users open the shopping app after the first recommendation.

[0102] The step of obtaining the user's favorite products and entering the favorite products into the desired product set includes:

[0103] Get all the items in the user's shopping cart and enter the desired item set;

[0104] The intended product collection is used to record the products that users have frequently browsed or have the intention to buy.

[0105] Set the quantity threshold a;

[0106] a is a constant, which can be adjusted according to actual conditions and is used to select a sufficient amount of historical data;

[0107] If the number of elements in the intended product set is less than a, the user's historical search keywords are obtained;

[0108] Get all products that match the historical search keywords and enter the intended product set;

[0109] By recording users' favorite products and building a collection of potential products, we can ensure that the recommended products meet users' preferences and increase users' click-through and purchase rates.

[0110] The dividing of time periods, obtaining historical data of the user in each time period, and calculating the average browsing time corresponding to each time period according to the historical data of each time period include:

[0111] Divide 24 hours into d time periods, and the length of each time period is equal;

[0112] Obtain the duration of each time the user uses the shopping app in history, and record it as browsing time;

[0113] Record browsing time into user browsing collection;

[0114] Divide all browsing times in the user browsing set into multiple sub-user browsing sets according to the time period at the start time;

[0115] By dividing the user's browsing behavior into multiple time periods and implementing different recommendation strategies for different time periods, we can capture the user's specific behavioral characteristics, thereby optimizing the recommendation effect and improving the relevance of recommended products and user satisfaction.

[0116] The start time is the start time of the browsing time;

[0117] Randomly select the sub-user browsing set as the target set;

[0118] Calculate the sum of all browsing time in the target set and record it as the total browsing time of the current target set;

[0119] Calculate the ratio between the total browsing time and the number of elements in the target collection;

[0120] By calculating this ratio, the average value of the user's historical browsing time in the time period of the target set can be calculated;

[0121] The ratio is recorded as the average browsing time of the time period to which the target set belongs;

[0122] Reselect the sub-user browsing set as the new target set, and calculate the average browsing time of the time period to which the target set belongs;

[0123] When all sub-user browsing sets are traversed, the selection calculation stops;

[0124] By utilizing historical data from different time periods and analyzing users’ browsing habits and the average browsing time for individual products, it helps to make personalized recommendations based on the user’s time period characteristics, thus increasing the effectiveness of the recommendations.

[0125] The method of obtaining the total number and total time of products browsed by the user in the current time period, calculating the user's single product browsing time, and determining the number n of product recommendations according to the single product browsing time and the average browsing time in the user's current time period includes:

[0126] If the user opens the shopping app, the time when the user opens the shopping app is obtained and recorded as the start time;

[0127] Determine the time period in which the start time falls, and obtain the sub-user browsing set corresponding to the time period;

[0128] Record the sub-user browsing set corresponding to the time period as the user monitoring set;

[0129] The purpose of setting the user monitoring set is to determine the sub-user browsing set corresponding to the time period when the current user opens the shopping software;

[0130] Get the total browsing time of the history corresponding to the user monitoring collection;

[0131] Calculate the sum of the number of historically browsed products at each browsing time in the user monitoring set, and record it as the total number of browses;

[0132] Calculate the ratio between the total browsing time of the user monitoring set and the total number of browsings, and record it as the browsing time of a single product;

[0133] This ratio refers to the average time that users have spent browsing a product in the history of the time period corresponding to the current user monitoring set;

[0134] At this time, the total browsing time and total number of browsing refer to historical data, and the information currently browsed by the user has not yet been entered into the collection;

[0135] The ratio of the average browsing time of the user monitoring set to the browsing time of a single product is recorded as the number of product recommendations n;

[0136] This ratio represents the number of items recommended to the user;

[0137] By randomly selecting products to predict user behavior and adjusting recommendation strategies, it is helpful to promptly correct recommended products that do not meet user interests, thereby improving the hit rate of product recommendations and user retention rate;

[0138] By determining in real time the time period in which the user opens the shopping software, obtaining the corresponding user historical browsing information based on the time period, calculating the user's browsing time for a single product in the current time period based on the historical browsing information, and recommending an appropriate number of products to the user based on the average browsing time and the browsing time for a single product in the current time period, and by obtaining the user's browsing situation in the current time period and making recommendations based on the current browsing time and average browsing time of the products, it can better adapt to the user's real-time needs and improve user experience and conversion rate.

[0139] The selecting of products according to the number of recommended products n and making the first recommendation, obtaining the actual number of products browsed by the user in the current actual usage time, and determining whether to enter the unbrowsed products in the recommended products into the pre-selected set include:

[0140] Randomly select a product from the set of n intended products as the predicted product;

[0141] Recommend the selected products to the user and remove the selected products from the intended product set;

[0142] Get the time interval of the user's current use of the shopping software and record it as the actual use time;

[0143] Get the number of products browsed by the user during the actual usage time, which is recorded as the actual number of products;

[0144] If the actual number of products is less than n, the unbrowsed products in the predicted products are marked and entered into the pre-selected set;

[0145] The elements in the pre-selected set are all the items that the user has not browsed in each recommended item;

[0146] If the actual number of products is greater than or equal to n, no entry operation is performed on the pre-selected set.

[0147] The updating of the user's relevant data according to the user's browsing of the recommended products includes:

[0148] The actual usage time is recorded in the sub-user browsing set of the time period;

[0149] Recalculate the sum of the elements in the current sub-user's browsing collection and use it as the total browsing time of the current sub-user's browsing collection;

[0150] Calculate the ratio between the total browsing time and the number of elements in the current sub-user's browsing set;

[0151] The ratio is recorded as the average browsing time of the time period to which the current sub-user browses the collection;

[0152] By updating and optimizing the user browsing behavior data and adjusting the recommendation model according to the latest browsing data each time, the accuracy of product recommendations can be continuously improved.

[0153] When the user opens the shopping software again, the product recommendation is performed according to the pre-selected set and the number m of products selected when the user opens the shopping software again, including:

[0154] If the user opens the shopping software again, the time when the user opens the shopping software is obtained and recorded as the first time;

[0155] Determine the time period of the first time, and obtain the sub-user browsing set corresponding to the time period;

[0156] Record the sub-user browsing set corresponding to the time period as the target set;

[0157] Get the total browsing time of the target collection;

[0158] Calculate the sum of the number of browsed products at each browsing time in the target set, and record it as the first value;

[0159] Calculate the ratio between the total browsing time of the target collection and the first value, and record it as the second value;

[0160] The ratio of the average browsing time of the target set to the second value is recorded as the number of selected products m;

[0161] The above content refers to the operation when a user makes a product recommendation again after having made a product recommendation. Although the two operations are the same, their meanings are different. The data referenced in the first product recommendation are all historical data. When making product recommendations at this time, the reference data is different because the historical data has been updated in the first product recommendation and a pre-selected set will be generated. Therefore, the reference data and methods for the two product recommendations are different.

[0162] If the number of elements in the pre-selected set is greater than or equal to m, the user's purchase amount each time is obtained and entered into the consumption set;

[0163] Get the maximum and minimum values ​​in the consumption set and record them as the maximum amount and minimum amount;

[0164] A user consumption capacity interval is formed according to the maximum amount and the minimum amount, specifically, the minimum amount is the minimum value of the user consumption capacity interval, and the maximum amount is the maximum value of the user consumption capacity interval;

[0165] Enter all elements in the pre-selected set whose prices are within the user's spending capacity into the preferred set;

[0166] The elements in the pre-selected set are all the products that the user has not browsed each time the product is recommended, while the elements in the preferred set are all the products in the pre-selected set whose prices are within the user's spending power range;

[0167] If the number of preferred sets is greater than or equal to m, randomly select m items from the preferred sets and enter them into the recommended item set, and recommend all items in the recommended item set to the user;

[0168] Eliminate the elements in the pre-selected set that are the same as the product recommendation set;

[0169] If the number of preferred sets is less than m, all elements in the preferred set are entered into the product recommendation set;

[0170] Calculate the difference between the number of optimal sets and m, and record it as the remaining number k;

[0171] Randomly select k items that are not selected in the pre-selected set and enter them into the product recommendation set;

[0172] Recommend all elements in the product recommendation set to the user;

[0173] Eliminate the elements in the pre-selected set that are the same as the product recommendation set;

[0174] By combining multi-dimensional information such as user browsing behavior, spending power, product prices, etc., we ensure that the recommended products not only meet the user's browsing habits, but also match the user's spending power range, thereby increasing the possibility of conversion.

[0175] When the user opens the shopping software again, the product recommendation is performed according to the pre-selected set and the number m of products selected when the user opens the shopping software again, including:

[0176] If the number of elements in the pre-selected set is less than m, all elements in the pre-selected set are entered into the product recommendation set;

[0177] Calculate the difference between m and the number of elements in the pre-selected set, and record the difference as c;

[0178] Enter all elements in the intended product set whose prices are within the user's consumption capacity range into the qualified product range;

[0179] Randomly select c elements from the qualified product set and enter them into the product recommendation set;

[0180] Recommend all elements in the product recommendation set to the user;

[0181] Eliminate the elements in the pre-selected set and the intended product set that are the same as the product recommendation set;

[0182] By obtaining the user's browsing history of recommended products, the unbrowsed products are entered into a pre-selected set for priority recommendation next time. When the number of elements in the pre-selected set is insufficient for the next recommendation, the user's consumption capacity range is obtained through the user's historical consumption record. Based on the user's consumption capacity range, products that meet the user's consumption capacity are recommended first. By adjusting the browsing history of recommended products, the recommended product set is continuously optimized to ensure that the number of products in the recommended product set is sufficient and meets the user's consumption capacity, thereby improving the accuracy of recommendations and the efficiency of product display.

Claims

1. A marketing method for online sales, characterized by: include, S1 obtains the user's favorite products and enters the favorite products into the intended product set; The step of obtaining the user's favorite products and entering the favorite products into the desired product set includes: Get all the items in the user's shopping cart and enter the desired item set; Set the quantity threshold a; If the number of elements in the intended product set is less than a, the user's historical search keywords are obtained; Get all products that match the historical search keywords and enter the intended product set; S2 divides the time period, obtains the historical data of the user in each time period, and calculates the average browsing time corresponding to each time period based on the historical data of each time period; The historical data specifically refers to the usage time from the start time to the end time each time the user uses the shopping software; The dividing of time periods, obtaining historical data of the user in each time period, and calculating the average browsing time corresponding to each time period according to the historical data of each time period include: Divide 24 hours into d time periods, and the length of each time period is equal; Obtain the duration of each time the user uses the shopping app in history, and record it as browsing time; Record browsing time into user browsing collection; Divide all browsing times in the user browsing set into multiple sub-user browsing sets according to the time period at the start time; The start time is the start time of the browsing time; Randomly select the sub-user browsing set as the target set; Calculate the sum of all browsing time in the target set and record it as the total browsing time of the current target set; Calculate the ratio between the total browsing time and the number of elements in the target collection; The ratio is recorded as the average browsing time of the time period to which the target set belongs; Reselect the sub-user browsing set as the new target set, and calculate the average browsing time of the time period to which the target set belongs; When all sub-user browsing sets are traversed, the selection calculation stops; S3: If the user opens the shopping software, the time period is determined according to the opening time, the total number and total time of the products browsed by the user in the current time period are obtained, and the browsing time of a single product by the user is calculated; Recommend products based on the user's browsing time and average browsing time for a single product in the current time period; The method of making product recommendations based on the browsing time and average browsing time of a single product in the current time period of the user includes: If the user opens the shopping app, the time when the user opens the shopping app is obtained and recorded as the start time; Determine the time period in which the start time falls, and obtain the sub-user browsing set corresponding to the time period; Record the sub-user browsing set corresponding to the time period as the user monitoring set; Get the total browsing time of the user monitoring collection; Calculate the sum of the number of browsed products at each browsing time in the user monitoring set, and record it as the total number of browsed products; Calculate the ratio between the total browsing time of the user monitoring set and the total number of browsings, and record it as the browsing time of a single product; The ratio of the average browsing time of the user monitoring set to the browsing time of a single product is recorded as the number of product recommendations n; Select products from n user-intended product sets to recommend to users and obtain user browsing information; S4 obtains the browsing status of the user on the recommended products, and adjusts the recommended products according to the browsing status; The obtaining of the browsing status of the user on the recommended products and adjusting the recommended products according to the browsing status includes: If the user opens the shopping software, the time when the user opens the shopping software is obtained and recorded as the first time; Determine the time period of the first time, and obtain the sub-user browsing set corresponding to the time period; Record the sub-user browsing set corresponding to the time period as the target set; Get the total browsing time of the target collection; Calculate the sum of the number of browsed products at each browsing time in the target set, and record it as the first value; Calculate the ratio between the total browsing time of the target collection and the first value, and record it as the second value; The ratio of the average browsing time of the target set to the second value is recorded as the number of selected products m; If the number of elements in the pre-selected set is greater than or equal to m, the user's purchase amount each time is obtained and entered into the consumption set; Get the maximum and minimum values ​​in the consumption set and record them as the maximum amount and minimum amount; A user consumption capacity interval is formed according to the maximum amount and the minimum amount, specifically, the minimum amount is the minimum value of the user consumption capacity interval, and the maximum amount is the maximum value of the user consumption capacity interval; Enter all elements in the pre-selected set whose prices are within the user's spending capacity into the preferred set; If the number of preferred sets is greater than or equal to m, randomly select m items from the preferred sets and enter them into the recommended item set, and recommend all items in the recommended item set to the user; Eliminate the elements in the pre-selected set that are the same as the product recommendation set; If the number of preferred sets is less than m, all elements in the preferred set are entered into the product recommendation set; Calculate the difference between the number of optimal sets and m, and record it as the remaining number k; Randomly select k items that are not selected in the pre-selected set and enter them into the product recommendation set; Recommend all elements in the product recommendation set to the user; Eliminate the elements in the pre-selected set that are the same as the product recommendation set; The obtaining of the browsing status of the user on the recommended products and adjusting the recommended products according to the browsing status includes: If the number of elements in the pre-selected set is less than m, all elements in the pre-selected set are entered into the product recommendation set; Calculate the difference between m and the number of elements in the pre-selected set, and record the difference as a; Enter all elements in the intended product set whose prices are within the user's consumption capacity range into the qualified product range; Randomly select a elements from the qualified product set and enter them into the product recommendation set; Recommend all elements in the product recommendation set to the user; Eliminate the elements in the pre-selected set and the intended product set that are the same as the product recommendation set; S5 updates the user's relevant data according to the user's browsing time and browsed products each time; Relevant data include the average browsing time and number of products browsed by users in different time periods; The updating of user-related data according to each user's browsing time and browsed products includes: The actual usage time is entered into the sub-user browsing time of the time period; Recalculate the sum of the elements in the current sub-user's browsing collection and use it as the total browsing time of the current sub-user's browsing collection; Calculate the ratio between the total browsing time and the number of elements in the current sub-user's browsing set; The ratio is recorded as the average browsing time of the current sub-user browsing collection in the time period.

2. A marketing method for online sales according to claim 1, characterized in that: The selecting of products from the n user intended product sets to recommend to the user and obtaining the user's browsing status includes: Randomly select a product from the set of n intended products as the predicted product; Recommend the selected products to the user and remove the selected products from the intended product set; Get the time interval of the user's current use of the shopping software and record it as the actual use time; Get the number of products browsed by the user during the actual usage time, which is recorded as the actual number of products; If the actual number of products is less than n, the unbrowsed products in the predicted products are marked and entered into the pre-selected set.

Images