E-commerce advertisement putting system and putting method based on genetic algorithm

Abstract

The invention discloses an e-commerce advertisement putting system and putting method based on a genetic algorithm, and the putting system comprises a data acquisition model and a genetic algorithm model. The delivery method comprises the following steps: generating a target e-commerce platform user data set; performing preprocessing and feature extraction through a statistical method; comparing the difference of the data set, and judging whether to output again; reading the feature data set of the target e-commerce platform, and outputting an advertisement putting strategy; reading an advertisement putting strategy, optimizing an advertisement putting scheme, and putting an advertisement; and evaluating an advertisement putting effect in real time, feeding back an evaluation result, and readjusting an advertisement putting strategy. According to the method, the genetic algorithm is adopted, the advertisement putting effect is monitored in real time through a multi-target optimization mechanism and a real-time monitoring feedback mechanism, the advertisement strategy and the putting content are adjusted in time by analyzing the specific behavior characteristic data feedback of the user, it is ensured that advertisement putting is always kept consistent with market changes and user requirements, and the user experience is improved. And the problem of user aversion caused by irrelevant advertisements is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of advertisement delivery, and in particular to an e-commerce advertisement delivery system and delivery method based on a genetic algorithm. Background Art

[0002] With the rapid development of digital marketing, e-commerce platforms have been facing a key challenge in the industry to attract potential customers through advertising strategies and improve sales conversion rates. As online shopping becomes a part of consumers' daily lives, how to stand out from the huge user base and improve the targeting and effectiveness of advertising has become an unavoidable key point for every e-commerce platform.

[0003] Traditional advertising methods usually rely on the experience and judgment of marketers and simple rule-based algorithms. They are difficult to adjust based on the immediate effects of advertising, ignore factors such as individual potential needs, emotional tendencies, and purchase decision cycles, and are unable to deeply mine potential information in user data, resulting in advertising content that cannot accurately match user needs and preferences, thereby affecting advertising reach and conversion rates, leading to a decline in user experience, and thus reducing the long-term effects and overall benefits of advertising. Summary of the invention

[0004] 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.

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an e-commerce advertising delivery system and delivery method based on genetic algorithm.

[0006] In order to achieve the above objectives, in a first aspect, the present invention provides an e-commerce advertising delivery system based on a genetic algorithm, comprising:

[0007] It includes a data acquisition model and a genetic algorithm model, wherein the data acquisition model is used to analyze user behavior data and adjust advertising delivery strategies, and the genetic algorithm model is used to read and analyze data and generate advertising delivery strategies;

[0008] The data acquisition model includes:

[0009] A data collection module, which is used to collect user behavior data and platform traffic data on the e-commerce platform;

[0010] A data processing module, which is used to pre-process and extract features of user behavior data in the data set by using statistical methods;

[0011] A data comparison module, which is used to compare the user behavior data before and after processing and determine whether to input into the genetic algorithm model;

[0012] An optimization data module, wherein the optimization data module is used to read the advertisement delivery strategy output by the genetic algorithm model and then adjust the advertisement delivery plan;

[0013] An advertisement delivery module, the advertisement delivery module is used to read the advertisement delivery plan output by the optimization data module and then deliver the advertisement;

[0014] The effect evaluation module is used to evaluate the effect of advertisement delivery and then feed back to the genetic algorithm model.

[0015] In some embodiments, the user behavior data includes browsing history, purchase history, search history, advertising display and click data.

[0016] In some embodiments, the platform traffic data includes time period traffic, and the time period traffic is collected by evenly dividing the time periods into 24 hours.

[0017] In some of the embodiments, the statistical method includes descriptive statistics and correlation analysis, and the statistical method is used in conjunction with time period traffic to analyze the tendency of user behavior data in each time period.

[0018] In some of the embodiments, the advertisement delivery plan includes a delivery platform, a delivery period, and delivery content.

[0019] In a second aspect, the present invention further provides an e-commerce advertising delivery method based on a genetic algorithm, which is executed by the e-commerce advertising delivery system based on a genetic algorithm as described in the first aspect, and the delivery method steps include:

[0020] S100, collecting user behavior data and platform traffic data of multiple e-commerce platforms to generate a user data set of a target e-commerce platform;

[0021] S200, preprocessing and feature extraction of the target e-commerce platform user data set by statistical methods, and outputting the target e-commerce platform feature data set;

[0022] S300, comparing the difference between the target e-commerce platform user data set and the target e-commerce platform feature data set, and determining whether to re-output the target e-commerce platform feature data set;

[0023] S400, reading the target e-commerce platform feature data set and outputting an advertising delivery strategy;

[0024] S500, reading the advertising delivery strategy, optimizing the advertising delivery plan and then delivering the advertisement;

[0025] S600, evaluates the advertising effect in real time, provides feedback on the evaluation results, and readjusts the advertising strategy.

[0026] In some embodiments, the step S200 includes:

[0027] S210, reading user behavior data and platform traffic data through descriptive statistics, pre-processing to generate statistical indicators, and then calculating values ​​through analytical methods;

[0028] S220, screening feature values ​​through correlation analysis, extracting feature values ​​to determine user inclination.

[0029] In some embodiments, the step S210 includes:

[0030] S211, reading user browsing history, purchase history, search history, advertisement display and click data;

[0031] S212, filtering and processing the traffic during the time period, and calculating the user behavior data and the conversion rate corresponding to the delivered advertisements.

[0032] In some embodiments, the step S220 includes:

[0033] S221, analyzing the user behavior data of different e-commerce platforms in each period and the conversion rate corresponding to the advertisements, and selecting highly relevant advertisements;

[0034] S222, analyzing the user's click and search preferences and evaluating the user's purchasing tendency.

[0035] In some embodiments, the step S300 includes:

[0036] S310, comparing the differences between the target e-commerce platform user dataset and the target e-commerce platform feature dataset;

[0037] S320, if there is a difference between the target e-commerce platform user dataset and the target e-commerce platform feature dataset, return to S310;

[0038] S330: If there is no difference between the target e-commerce platform user data set and the target e-commerce platform feature data set, proceed to S400.

[0039] The present invention has the following beneficial effects:

[0040] 1. The present invention adopts a genetic algorithm, combines the user behavior data and traffic data of the e-commerce platform, and accurately identifies the peak period of user traffic through efficient global search capabilities. It can capture traffic fluctuations in real time and accurately locate the peak period of user traffic. The algorithm can not only make predictions based on historical behavior data, but also flexibly respond to real-time traffic changes, ensuring that the identification of peak periods is timely and accurate;

[0041] 2. It adopts a multi-objective optimization mechanism and a real-time monitoring feedback mechanism, which can monitor the effect of advertising in real time, and adjust the advertising strategy and delivery content in a timely manner through data feedback, ensuring that advertising is always highly consistent with market changes and user needs, improving the relevance of advertising content and user experience, and enabling advertising to maximize the advertising effect without sacrificing user experience, thus avoiding user aversion caused by irrelevant advertising. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a principle framework diagram of the e-commerce advertising delivery system based on genetic algorithm proposed by the present invention;

[0043] Figure 2 The process of the e-commerce advertising delivery system based on genetic algorithm proposed by the present invention Figure 1 ;

[0044] Figure 3 The process of the e-commerce advertising delivery system based on genetic algorithm proposed by the present invention Figure 2 ;

[0045] Figure 4 The process of the e-commerce advertising delivery system based on genetic algorithm proposed by the present invention Figure 3 ;

[0046] Figure 5 The process of the e-commerce advertising delivery system based on genetic algorithm proposed by the present invention Figure 4 .

[0047] Legend:

[0048] 1. Data acquisition model; 101. Data collection module; 102. Data processing module; 103. Data comparison module; 104. Data optimization module; 105. Advertising delivery module; 106. Effect evaluation module; 2. Genetic algorithm model. DETAILED DESCRIPTION

[0049] 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.

[0050] The embodiments of the present application provide an e-commerce advertising delivery system and delivery method based on a genetic algorithm, which solves the problem that the traditional advertising delivery method in the prior art usually relies on the experience and judgment of marketers and simple rule algorithms, is difficult to adjust according to the immediate effect of advertising delivery, ignores factors such as individual potential needs, emotional tendencies, and purchase decision cycles, and is unable to deeply mine potential information in user data, resulting in the inability to accurately match advertising content with user needs and preferences, thereby affecting the reach and conversion rate of advertising, resulting in a decline in user experience, thereby reducing the long-term effect and comprehensive benefits of advertising delivery. This application combines the user behavior data and traffic data of the e-commerce platform, and through efficient global search capabilities, it can accurately identify the peak hours of user traffic, discover the patterns of user activity, capture traffic fluctuations in real time, and accurately locate the peak hours of user traffic. The algorithm can not only make predictions based on historical behavior data, but also flexibly respond to real-time traffic changes to ensure that the identification of peak hours is timely and accurate; it adopts a multi-objective optimization mechanism and a real-time monitoring feedback mechanism, which can monitor the effectiveness of advertising in real time, and adjust advertising strategies and delivery content in a timely manner through data feedback to ensure that advertising is always highly consistent with market changes and user needs, and ensure the relevance of advertising content and user experience, so that advertising can maximize the advertising effect without sacrificing user experience, and avoid user aversion caused by irrelevant advertising.

[0051] Please refer to the following examples for details:

[0052] Reference Figure 1 The present invention provides an embodiment of an e-commerce advertising delivery system based on a genetic algorithm, including a data acquisition model and a genetic algorithm model. The data acquisition model is used to analyze user behavior data and adjust advertising delivery strategies, and the genetic algorithm model is used to read and analyze data and generate advertising delivery strategies.

[0053] The data acquisition model includes:

[0054] (1) Data collection module, which collects rich and diverse user behavior data and platform traffic data on the e-commerce platform. User behavior data covers a lot of user operation information on the platform, such as detailed browsing records of users, including the types of products browsed, browsing time, browsing order, and the time spent on each page; in terms of purchase records, key elements such as the name, quantity, price, purchase time, and purchase frequency of the products purchased by users will be recorded; search records will also be collected in detail, such as search keywords, number of searches, search time points, and browsing conditions of search results; at the same time, advertising display data will be collected, such as the location of advertising display, number of displays, and which specific user groups are displayed to, as well as advertising click data, such as the time of click, number of clicks, and the subsequent behavior path of users after clicking the advertisement. Platform traffic data focuses on the traffic scale of the platform in different time periods. According to specific time division standards, such as every hour or every half hour as a statistical interval, accurate statistics are made on the number of users entering the platform, the total number of pages browsed, and the traffic distribution of each page, so as to provide sufficient and accurate data support for subsequent analysis and decision-making.

[0055] (2) Data processing module, using statistical methods to deeply process the user behavior data in the data set. In the preprocessing stage, the collected data will be cleaned to remove possible erroneous data, duplicate data, and incomplete data to ensure the quality and reliability of the data. Feature extraction is then performed through statistical calculations. For example, for browsing records, the average browsing time of a specific product or page, the distribution of browsing times, and the fluctuation trend of browsing volume in different time periods are calculated; for purchase records, the mean, median, and mode of the purchase amount, the distribution of the number of purchased products, and the difference characteristics of purchase behavior of different user groups are analyzed; in terms of search records, the frequency of occurrence of popular search keywords, the correlation characteristics between search keywords and purchase behavior, etc. are extracted; for advertising data, the conversion rate characteristics between advertising display and clicks, the effect characteristics of different advertising positions, etc. are statistically analyzed, so as to transform the original user behavior data into feature data with high analytical value and decision-making reference, so as to better provide accurate data sets for subsequent modules.

[0056] (3) The data comparison module conducts a detailed comparative analysis of the user behavior data before and after being processed by the data processing module, and makes judgments from multiple dimensions such as data integrity, accuracy, and consistency. For example, it compares whether the number of users in the data before and after processing matches, whether the value range of each data field is reasonable, and whether the logical relationship of the data before and after processing is consistent. Only when the data is determined to fully meet the requirements and there are no abnormalities will it be input into the genetic algorithm model for further calculation and optimization processing. If any problems are found in the data, they will be promptly fed back to the data processing module for reprocessing or correction.

[0057] (4) After receiving the advertising delivery strategy output by the genetic algorithm model, the optimization data module will make comprehensive adjustments and optimizations to the advertising delivery plan based on factors such as the actual operation of the platform, market dynamics, and real-time fluctuations in user demand. For example, based on the current distribution of peak and trough traffic periods on the platform, the timing of advertising delivery will be finely adjusted to focus advertising delivery during peak traffic periods and when users have a high purchasing intention, while avoiding trough traffic periods or periods with low user activity; based on the characteristics and behavioral preferences of different user groups, the target audience for advertising delivery will be more accurately positioned and segmented to ensure that the advertisement can accurately reach the user group most likely to make a purchase; with regard to the content form of advertising delivery, based on market feedback and user preferences, the most attractive and influential advertising ideas, copywriting, pictures or videos and other elements will be selected for combination and optimization, so as to develop a set of advertising delivery plans that can achieve the best results in the actual delivery environment, so as to maximize the efficiency and effectiveness of advertising delivery.

[0058] (5) Advertisement delivery module, which delivers advertisements according to delivery parameters such as delivery time, delivery platform location, and delivery target user group specified in the advertisement delivery plan. For example, some high-priority advertisements will be displayed on key locations such as the platform's homepage, popular product page, or user personalized recommendation page, and will be rotated at a higher frequency; and for advertisements for some specific audience groups, the platform's user portrait and precision delivery system will accurately deliver the advertisements to eligible users, ensuring the accuracy and effectiveness of the advertisement delivery, so that the advertisements can be displayed at the right time, in the right place, and in front of the right audience, thereby stimulating users' purchasing interest and desire and achieving the commercial goals of the advertisement delivery;

[0059] (6) The effect evaluation module collects the click-through rate and conversion rate of the advertisements, evaluates the ratio of the number of times the advertisements are clicked by users to the number of times the advertisements are displayed, and the ratio of the number of users who actually complete the purchase behavior after clicking the advertisement to the number of users who click the advertisement. The evaluation results are then fed back to the genetic algorithm model in a timely manner so that the genetic algorithm model can further optimize and adjust the advertising delivery strategy based on this feedback information, so that the entire advertising delivery system can form a closed-loop process of continuous self-improvement and self-optimization, and continuously improve the effect and quality of advertising delivery.

[0060] Furthermore, user behavior data includes:

[0061] (1) Browsing history: a detailed record of every product page, store page, and various topic pages that a user has browsed on an e-commerce platform, including the time period of browsing, the length of time spent on a page, the order in which pages were browsed, and whether the page was scrolled, zoomed in, or otherwise;

[0062] (2) Purchase records: basic information such as name, brand, model, specification, quantity, unit price and total price of the purchased goods, purchase time, payment method, store information, etc.;

[0063] (3) Search history: the keywords entered by the user in the platform search box, the time and number of searches, and the browsing status of the search results page, and whether filtering conditions were used, such as price range, brand filtering, color filtering, etc.;

[0064] (4) Advertising display: the location of the advertisement on the platform (such as the carousel advertisement on the homepage, the sidebar advertisement on the product details page, the top advertisement on the search results page, etc.), the time of display, the specific user groups to which the advertisement is displayed (determined based on user profile information), and the type of advertisement (image advertisement, video advertisement, text link advertisement, etc.);

[0065] (5) Click data: information such as the time when a user clicks on the ad, the number of clicks, the page to which the user is redirected after clicking the ad, the length of time the user stays on the page, and subsequent operations.

[0066] Furthermore, platform traffic data includes time period traffic, which is collected evenly in 24 hours. A day can be divided into 12 time periods, each lasting 2 hours. For example, the first time period is 0:00-2:00, the second time period is 2:00-4:00, and so on. In each time period, accurate statistics of key traffic indicators such as the number of unique visitors (UV), page views (PV), visit depth (the number of pages viewed by a user in one visit), and average visit duration entering the e-commerce platform will provide a basic data set for subsequent analysis of user behavior in different time periods;

[0067] It should be explained in detail that the advertising plan includes the delivery platform, delivery time period and delivery content. The statistical methods include descriptive statistics and correlation analysis. The statistical methods are combined with the time period traffic to analyze the tendency of user behavior data in each time period:

[0068] (1) Descriptive statistics:

[0069] Browsing record analysis: In addition to the regular page views (PV), it also calculates the unique visitor views (UV_PV, that is, the number of pages viewed by each unique visitor), page browsing time distribution (statistics of the proportion of page views in different time intervals, such as 0-30 seconds, 30 seconds-1 minute, 1-3 minutes, etc.), popular browsing page rankings (pages are sorted according to the number of views to determine which pages are most popular with users), etc.; statistics on daily, weekly or monthly page views change trends with timestamps accurate to milliseconds, draw page views time series graphs, and analyze users' active time patterns in different time periods (such as weekdays and weekends, daytime and nighttime, etc.).

[0070] Purchase record analysis: In addition to basic indicators such as the number of purchases, the number of purchased items, and the purchase amount, we also calculate the average order value (the average amount of each order), the distribution of purchased item types (the proportion of purchases of different types of items), the repurchase rate (the ratio of users who have purchased more than 1 times to the total number of purchasing users), etc.; calculate the conversion rate (the ratio of the number of purchasers to the number of browsers), and conduct conversion rate analysis based on dimensions such as different product categories, different time periods, and different user groups (such as new users and old users) to deeply explore the characteristics of user purchasing behavior.

[0071] Search record analysis: Based on search keywords, search times, search result click-through rate (the ratio of the number of times a search result is clicked to the number of times a search result is displayed), etc., statistics are also collected on the popularity change trend of popular search keywords (analyzing the change in the number of keyword search times through time series), the long-tail distribution of search keywords (counting the number and proportion of low-frequency but specifically targeted search keywords), etc.; analysis of changing trends in user search habits, such as the amount of time in advance that users search for keywords for seasonal products (e.g. how long before winter comes to start searching for warm products such as down jackets), and optimization of the search function based on the correlation between search keywords and product categories to improve the accuracy of search results.

[0072] Ad display and click data analysis: In addition to basic indicators such as display volume, click volume, click-through rate (ratio of click volume to display volume), there are also calculations of ad exposure cost (ratio of ad placement cost to display volume), ad click cost (ratio of ad placement cost to click volume), ad conversion rate (ratio of the number of people who purchase due to ad clicks to ad click volume), etc.; calculate click-through rate, and analyze click-through rate according to dimensions such as different ad positions, different time periods, different ad types, and different user groups, to evaluate advertising effectiveness and optimize advertising strategies.

[0073] (2) Correlation analysis

[0074] Browsing and purchasing correlation analysis: Use data mining algorithms such as association rule mining to deeply analyze the correlation between browsing times and purchasing behavior;

[0075] Search and purchase correlation analysis: Analyze the correlation between search keywords and purchase behavior to explore user purchase preferences. By establishing a mapping relationship between search keywords and purchase product categories, determine which search keywords are highly correlated with which product purchase behaviors.

[0076] Advertising data correlation analysis: Use correlation analysis to analyze the relationship between data such as advertising display volume and click volume and purchasing behavior to evaluate advertising effectiveness; at the same time, analyze the correlation between advertising click volume and purchase amount to understand which advertising clicks can bring higher purchase amounts, thereby optimizing advertising delivery strategies and improving the return on investment of advertising delivery.

[0077] Analysis of correlation between display and click: Analyze the correlation between ad display and click to determine the attractiveness of the ad. In addition to calculating the click-through rate, you can also analyze the changes in click-through rate within different ad display ranges to determine the amount of ad display at which the click-through rate will stabilize or begin to decline, so as to adjust the frequency and location of ad delivery.

[0078] Click-to-conversion correlation analysis: Analyze the correlation between ad clicks and conversion rates, optimize advertising strategies, and determine which ad clicks can be effectively converted into purchases by analyzing the changes in conversion rates at different ad click levels, and which ad clicks have high click volumes but low conversion rates and need to be optimized.

[0079] Furthermore, the genetic algorithm model uses user behavior data and time period traffic:

[0080] (1) Initializing the population: Generate an initial population based on user behavior data and traffic data. Each individual represents a combination of advertising delivery strategies (including advertising content, delivery time period, etc.);

[0081] (2) Fitness function: The fitness function evaluates the quality of individuals based on indicators such as the click-through rate, conversion rate, and user activity of the advertisement. Individuals with higher fitness values ​​have more effective advertising strategies.

[0082] (3) Selection, crossover, and mutation: Through the selection operation, individuals with high fitness are selected from the current population, and crossover and mutation operations are performed to generate new advertising delivery strategies. Crossover and mutation operations can generate more diverse strategy combinations and avoid local optimal solutions.

[0083] (4) Genetic operation iteration: Continuously iterate the genetic algorithm until the optimal advertising delivery strategy is found.

[0084] (5) Output the optimal delivery strategy: Finally, the optimal advertising delivery strategy is output and handed over to the advertising delivery module for execution.

[0085] Reference Figure 2-Figure 5 The present invention also provides an embodiment of an e-commerce advertising delivery method based on a genetic algorithm, which is executed by an e-commerce advertising delivery system based on a genetic algorithm as in the above embodiment, and the delivery method steps include:

[0086] S100, collecting user behavior data and platform traffic data of multiple e-commerce platforms to generate a user data set of a target e-commerce platform;

[0087] Comprehensively collect user behavior data and platform traffic data from multiple e-commerce platforms, integrate the data scattered across various platforms through efficient data collection technology and interfaces, and generate a user data set for the target e-commerce platform. This data set covers a massive amount of user behavior information and traffic fluctuation data, providing a basis for subsequent analysis.

[0088] S200, preprocessing and feature extraction of the target e-commerce platform user data set by statistical methods, and outputting the target e-commerce platform feature data set;

[0089] Please continue reading Figure 3 In this embodiment, step S200 includes:

[0090] S210, reading user behavior data and platform traffic data through descriptive statistics, pre-processing to generate statistical indicators, and then calculating values ​​through analytical methods;

[0091] S220, screening feature values ​​through correlation analysis, extracting feature values ​​to determine user inclination;

[0092] Please continue reading Figure 4 In this embodiment, step S210 includes:

[0093] S211, reading user browsing history, purchase history, search history, advertisement display and click data;

[0094] S212, filtering the traffic during the processing period, calculating the user behavior data and the conversion rate corresponding to the advertisements placed;

[0095] Please continue reading Figure 4 In this embodiment, step S220 includes:

[0096] S221, analyzing the user behavior data of different e-commerce platforms in each period and the conversion rate corresponding to the advertisements, and selecting highly relevant advertisements;

[0097] S222, analyzing the user's click and search preferences and evaluating the user's purchasing tendency;

[0098] S300, comparing the difference between the target e-commerce platform user data set and the target e-commerce platform feature data set, and determining whether to re-output the target e-commerce platform feature data set;

[0099] Please continue reading Figure 5 In this embodiment, step S300 includes:

[0100] S310, comparing the differences between the target e-commerce platform user dataset and the target e-commerce platform feature dataset;

[0101] S320, if there is a difference between the target e-commerce platform user dataset and the target e-commerce platform feature dataset, return to S310;

[0102] S330, if there is no difference between the target e-commerce platform user data set and the target e-commerce platform feature data set, proceed to S400;

[0103] S400, reading the target e-commerce platform feature data set and outputting an advertising delivery strategy;

[0104] S500, reading the advertising delivery strategy, optimizing the advertising delivery plan and then delivering the advertisement;

[0105] S600, evaluates the advertising effect in real time, provides feedback on the evaluation results, and readjusts the advertising strategy.

[0106] Accurately read detailed information such as user browsing history, purchase history, search history, ad display and click data, and timely filter and process traffic during the processing period, and then calculate user behavior data and the conversion rate corresponding to the advertising according to different data types. For example, when calculating the purchase conversion rate, the number of purchases is divided by the number of views or ad clicks to quantitatively evaluate the effect of advertising on purchasing behavior. Through these operations, a series of representative statistical indicators are generated, such as page views PV, purchases, searches, etc., and analytical methods are used to deeply explore the user behavior patterns behind these indicators.

[0107] In-depth analysis of user behavior data and the conversion rate of advertisements in each period of different e-commerce platforms, and selection of highly relevant advertisements through correlation algorithms. For example, the correlation between different variables is calculated using methods such as the Pearson correlation coefficient to find out the types of advertisements and user behavior characteristics that are highly correlated with purchase behavior; at the same time, a detailed analysis of user click search preferences is conducted, and through a comprehensive analysis of search keywords, search frequency, search result click distribution and other data, user purchase tendencies are evaluated, and characteristic values ​​that have a key impact on the formulation of advertising delivery strategies are extracted, thereby constructing a characteristic data set for the target e-commerce platform.

[0108] Evaluate the advertising effect in real time, and use a multi-dimensional evaluation index system, such as advertising exposure, click-through rate, conversion rate, user retention rate, etc., to conduct comprehensive and real-time monitoring and evaluation of the advertising effect; then feedback the evaluation results in a timely manner, and readjust the advertising strategy based on the feedback information. For example, if it is found that the conversion rate of an advertisement on a specific platform or time period is low, the target audience positioning of the advertisement may be adjusted, the creative content of the advertisement may be optimized, or the delivery platform and time period may be changed to further improve the efficiency and effect of advertising, achieve continuous optimization and improvement of the advertising effect, and ensure that advertising always adapts to market changes and user needs.

[0109] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An e-commerce advertising delivery system based on genetic algorithm, characterized in that: It includes a data acquisition model and a genetic algorithm model, wherein the data acquisition model is used to analyze user behavior data and adjust advertising delivery strategies, and the genetic algorithm model is used to read and analyze data and generate advertising delivery strategies; The data acquisition model includes: A data collection module, which is used to collect user behavior data and platform traffic data on the e-commerce platform; A data processing module, which is used to pre-process and extract features of user behavior data in the data set by using statistical methods; A data comparison module, which is used to compare the user behavior data before and after processing and determine whether to input into the genetic algorithm model; An optimization data module, wherein the optimization data module is used to read the advertisement delivery strategy output by the genetic algorithm model and then adjust the advertisement delivery plan; An advertisement delivery module, the advertisement delivery module is used to read the advertisement delivery plan output by the optimization data module and then deliver the advertisement; The effect evaluation module is used to evaluate the effect of advertisement delivery and then feed back to the genetic algorithm model.

2. The e-commerce advertising delivery system based on genetic algorithm according to claim 1, characterized in that: The user behavior data includes browsing history, purchase history, search history, advertising display and click data.

3. The e-commerce advertising delivery system based on genetic algorithm according to claim 1, characterized in that: The platform traffic data includes time period traffic, and the time period traffic is collected according to the 24-hour average time period.

4. The e-commerce advertising delivery system based on genetic algorithm according to claim 1, characterized in that: The statistical method includes descriptive statistics and correlation analysis, and the statistical method is used in conjunction with the time period traffic to analyze the tendency of user behavior data in each time period.

5. The e-commerce advertising delivery system based on genetic algorithm according to claim 1, characterized in that: The advertisement delivery plan includes a delivery platform, a delivery period, and delivery content.

6. An e-commerce advertising method based on genetic algorithm, characterized in that: The delivery method is executed via the e-commerce advertising delivery system based on a genetic algorithm according to any one of claims 1 to 5, and the delivery method steps include: S100, collecting user behavior data and platform traffic data of multiple e-commerce platforms to generate a user data set of a target e-commerce platform; S200, preprocessing and feature extraction of the target e-commerce platform user data set by statistical methods, and outputting the target e-commerce platform feature data set; S300, comparing the difference between the target e-commerce platform user data set and the target e-commerce platform feature data set, and determining whether to re-output the target e-commerce platform feature data set; S400, reading the target e-commerce platform feature data set and outputting an advertising delivery strategy; S500, reading the advertising delivery strategy, optimizing the advertising delivery plan and then delivering the advertisement; S600, real-time evaluation of advertising effectiveness, feedback of evaluation results, and readjustment of advertising strategies.

7. The e-commerce advertising method based on genetic algorithm according to claim 6, characterized in that: The step S200 includes: S210, reading user behavior data and platform traffic data through descriptive statistics, pre-processing to generate statistical indicators, and then calculating values ​​through analytical methods; S220, screening feature values ​​through correlation analysis, extracting feature values ​​to determine user inclination.

8. The e-commerce advertising method based on genetic algorithm according to claim 7, characterized in that: The step S210 includes: S211, reading user browsing history, purchase history, search history, advertisement display and click data; S212, filtering and processing the traffic during the time period, and calculating the user behavior data and the conversion rate corresponding to the delivered advertisements.

9. The e-commerce advertising method based on genetic algorithm according to claim 7, characterized in that: The step S220 includes: S221, analyzing the user behavior data of different e-commerce platforms in each period and the conversion rate corresponding to the advertisements, and selecting highly relevant advertisements; S222, analyzing the user's click and search preferences and evaluating the user's purchasing tendency.

10. The e-commerce advertising delivery method based on genetic algorithm according to claim 6, characterized in that: The step S300 includes: S310, comparing the differences between the target e-commerce platform user dataset and the target e-commerce platform feature dataset; S320, if there is a difference between the target e-commerce platform user dataset and the target e-commerce platform feature dataset, return to S310; S330: If there is no difference between the target e-commerce platform user data set and the target e-commerce platform feature data set, proceed to S400.

Images