Advertisement pushing method, server, storage medium and program product

Abstract

An advertisement pushing method, a server, a storage medium and a program product relate to the field of advertisement pushing, and the method comprises the following steps: based on a preset advertisement database, determining a plurality of content elements in an advertisement and proportions respectively corresponding to the plurality of content elements; obtaining the staying duration of the user on the plurality of advertisements; determining the advertisements with the staying duration exceeding the preset duration as possible interested advertisements; determining interested content elements of the user according to the possibly interested content elements in each possibly interested advertisement and the proportions corresponding to the possibly interested content elements; and determining the advertisement of which the total proportion of the interested content elements exceeds a preset total proportion threshold as a recommended advertisement, and pushing the recommended advertisement to the user. By implementing the method, the accuracy of advertisement pushing can be improved.

Description

Technical Field

[0001] The present application relates to the field of advertisement push, and in particular to an advertisement push method, a server, a storage medium and a program product. Background Art

[0002] In the field of digital marketing, the effectiveness of advertising push is the key to improving user engagement and increasing advertising revenue. With the popularization of the Internet and the increase in online consumption behavior, companies are increasingly relying on precise advertising technology to attract target users. Therefore, how to accurately identify user interests and push relevant advertisements accordingly has become a current research hotspot.

[0003] In the related technologies, most advertising push systems usually use basic user behavior tracking technology to push advertising content. For example, user data is collected by tracking user activities on the Internet (such as browsing history, click records, etc.), and then these user data are used to build user portraits, thereby inferring user interests and selecting corresponding advertising content for delivery.

[0004] Although related technologies can infer user interests based on user historical data, this method often finds it difficult to accurately capture changes in user interests or respond to new points of interest in real time, because user interests may change with time, environment, or mood, and analysis based on user historical data may not be sufficient to capture these dynamic changes. Summary of the invention

[0005] The present application provides an advertisement push method, a server, a storage medium and a program product for improving the accuracy of advertisement push.

[0006] In a first aspect, the present application provides an advertisement push method, which is applied to a server, and the method includes: based on a preset advertisement database, determining multiple content elements in an advertisement and the respective corresponding proportions of the multiple content elements; obtaining the user's stay time on multiple advertisements; determining the advertisement whose stay time exceeds the preset time as an advertisement of possible interest; determining the user's content elements of interest based on the possible content elements of interest in each of the advertisements of possible interest and the corresponding proportions of the possible content elements of interest; determining the advertisement whose total proportion of the content elements of interest exceeds a preset total proportion threshold as a recommended advertisement, and pushing the recommended advertisement to the user.

[0007] By adopting the above technical solution, the server determines the user's potentially interesting advertisements based on the user's stay time on multiple advertisements. Since each potentially interesting advertisement includes multiple potentially interesting content elements and the corresponding proportions of multiple potentially interesting content elements, the server can determine the user's real interesting content elements by analyzing the user's potentially interesting advertisements. Based on these interesting content elements, the user's interesting advertisements are determined more accurately, and advertisements whose total proportion of interesting content elements exceeds the preset total proportion threshold are used as recommended advertisements and pushed to the user, achieving more accurate advertisement push, thereby increasing the probability of users clicking on advertisements, increasing the conversion rate of advertisements, and avoiding the waste of advertising resources. At the same time, this method also reduces the user's boredom with irrelevant advertisements and improves the user's experience in browsing advertisements.

[0008] In combination with some embodiments of the first aspect, in some embodiments, the determining of the user's content elements of interest based on the possible content elements of interest in each of the possible advertisements of interest and the corresponding ratios of the possible content elements of interest specifically includes: calculating the element ratio of each of the possible content elements of interest in a set of possible advertisements of interest based on the possible content elements of interest in each of the possible advertisements of interest and the corresponding ratios of the possible content elements of interest, the set of possible advertisements of interest including all possible advertisements of interest; determining the possible content elements of interest whose element ratio exceeds a preset element ratio threshold as the user's content elements of interest.

[0009] By adopting the above technical solution, the server calculates the element ratio of each potentially interesting content element in the potentially interesting advertisement set, and sets a preset element ratio threshold, so as to more accurately determine the user's content elements of interest. This calculation method takes into account the overall performance of the potentially interesting content elements in multiple potentially interesting advertisements, rather than relying solely on a single potentially interesting advertisement, and can therefore more comprehensively reflect user interests. By setting a preset element ratio threshold, elements with a low element ratio can be filtered out, and only the truly important content elements of interest are retained, which not only improves the accuracy of the content elements of interest, but also reduces the noise in subsequent advertising pushes, so that the recommended advertisements finally pushed are more in line with user interests, thereby improving the click-through rate and conversion rate of advertisements.

[0010] In combination with some embodiments of the first aspect, in some embodiments, determining that an advertisement whose total proportion of the content elements of interest exceeds a preset total proportion threshold is a recommended advertisement, and pushing the recommended advertisement to the user, specifically includes: calculating the total proportion of the content elements of interest in each advertisement in the preset advertisement database; determining the advertisement whose total proportion of the content elements of interest is higher than the preset ratio as the recommended advertisement; and pushing the recommended advertisements to the user in order of the total proportion of the content elements of interest.

[0011] By adopting the above technical solution, the server calculates the total proportion of the content elements of interest of each advertisement in the preset advertisement database, and sets a preset ratio to filter the recommended advertisements, thereby achieving more personalized and accurate advertisement push. At the same time, this method also avoids the push of advertisements with a single content element, thereby improving the user's acceptance and participation in the advertisement, enhancing the user experience, and ultimately maximizing the advertising effect.

[0012] In combination with some embodiments of the first aspect, in some embodiments, before the step of obtaining the length of time the user stays on multiple advertisements, the method also includes: displaying verification information; receiving the user's operation response to the verification information; if the operation response is correct, continuing to play the advertisement; if the operation response is incorrect, stopping playing the advertisement.

[0013] By adopting the above technical solution, the server displays verification information before playing the advertisement and requires the user to respond to the verification information, which effectively improves the accuracy and reliability of the advertising data. This verification mechanism can first ensure that the user is actually watching the advertisement, rather than leaving the device in an unattended state, thereby avoiding the collection of false dwelling time. Secondly, it can screen out users who are truly focused on the advertising content and improve the quality of subsequent data analysis. Furthermore, it can enhance the user's attention and engagement, and increase the user's memory and understanding of the advertising content. Finally, the mechanism of continuing to play the advertisement with a correct response and stopping the advertisement with an incorrect response not only respects the user's right to choose, but also optimizes the allocation efficiency of advertising resources.

[0014] In combination with some embodiments of the first aspect, in some embodiments, after the step of determining that an advertisement whose total proportion of the content elements of interest exceeds a preset total proportion threshold is a recommended advertisement, and pushing the recommended advertisement to the user, the method further includes: obtaining the facial expression and eye gaze duration of the user when watching the recommended advertisement; determining the user's satisfaction with the recommended advertisement based on the facial expression and the eye gaze duration; and when the satisfaction score is higher than the preset score threshold, increasing the push frequency of the recommended advertisement.

[0015] By adopting the above technical solution, the server obtains the facial expressions and eye gaze duration of users when watching recommended ads, achieving a deeper analysis of user interests and more accurate advertising delivery. First, the analysis of facial expressions can capture the user's emotional response to the advertising content, such as joy, boredom, etc., providing richer user feedback information than the dwell time. Secondly, the monitoring of eye gaze duration can accurately locate the content elements that users pay attention to, which helps to understand which content elements are most attractive to users. By comprehensively evaluating user satisfaction through these two indicators, the advertising effect can be judged more accurately. When the satisfaction score is higher than the preset score threshold, the push frequency of recommended ads is increased, which can not only strengthen the advertising content that users like, but also improve the efficiency and accuracy of overall advertising delivery.

[0016] In combination with some embodiments of the first aspect, in some embodiments, after the step of determining the user's satisfaction with the recommended advertisement based on the facial expression and the eye gaze duration, the method further includes: when the satisfaction score is lower than a preset score threshold, obtaining the reason for the user's dissatisfaction; and adjusting the advertising push strategy based on the reason for dissatisfaction.

[0017] By adopting the above technical solution, when the satisfaction score is lower than the preset score threshold, the server obtains the reason for the user's dissatisfaction and adjusts the advertising push strategy accordingly, thus achieving self-optimization and continuous improvement of the advertising push method. This method can not only continuously optimize the advertising push for a single user, but also discover and solve common problems in advertising push by summarizing and analyzing feedback from multiple users, thereby comprehensively improving the quality and effect of advertising push.

[0018] In combination with some embodiments of the first aspect, in some embodiments, the method also includes: obtaining characteristic information of the user, the characteristic information including gender, age and dress; predicting the personalized needs of the user based on the gender, age and dress; determining in the preset advertising database a personalized advertisement whose relevance to the personalized need exceeds a preset relevance threshold, and pushing the personalized advertisement to the user.

[0019] By adopting the above technical solution, the server obtains the user's characteristic information (including gender, age and dress), predicts the user's personalized needs, and pushes highly relevant personalized advertisements, thereby achieving more intelligent and accurate advertising push. First, based on the user's characteristic information, the user can be quickly profiled, and relatively accurate advertising recommendations can be provided even for new users. Secondly, by analyzing the user's dress, deeper information such as the user's lifestyle and aesthetic preferences can be inferred, which helps to more accurately predict the user's personalized needs. Furthermore, personalized advertisements with high relevance are screened in the preset advertising database to ensure the matching degree between advertising push and user's personalized needs. This method can not only improve the pertinence and effectiveness of advertising, but also provide a good advertising experience when the user first contacts it, which is conducive to improving the user's acceptance and participation in advertising, thereby improving advertising conversion rate and user satisfaction.

[0020] In a second aspect, an embodiment of the present application provides a server, comprising: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the server to execute the method described in the first aspect and any possible implementation method of the first aspect.

[0021] In a third aspect, an embodiment of the present application provides a computer program product comprising instructions, which, when executed on a server, enables the server to execute the method described in the first aspect and any possible implementation of the first aspect.

[0022] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, comprising instructions, which, when executed on a server, enable the server to execute the method described in the first aspect and any possible implementation of the first aspect.

[0023] It is understandable that the server provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the method provided in the embodiment of the present application. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding method, which will not be repeated here.

[0024] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. Due to the adoption of dynamic interest recognition and precise advertising push methods based on real-time user behavior, it is able to accurately capture real-time changes in user interests and respond quickly, effectively solving the lag and inaccuracy problems caused by relying solely on historical data in related technologies, thereby achieving highly personalized and real-time optimization of advertising push.

[0025] 2. Due to the adoption of a multi-dimensional user feedback collection and analysis mechanism, it is possible to comprehensively evaluate advertising effects and continuously optimize advertising push strategies, effectively solving the problems of single user feedback and inaccurate advertising effect evaluation in related technologies, thereby achieving self-learning and continuous improvement of advertising push.

[0026] 3. Due to the use of intelligent prediction and personalized recommendation technology based on user feature information, it is possible to provide high-quality advertising services when users first come into contact with the service, effectively solving the cold start problem of new users and the poor generalization of advertising recommendations in related technologies, thereby achieving rapid, accurate and wide applicability of advertising push. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a flow chart of the advertisement push method in the embodiment of the present application; Figure 2 is another flowchart of the advertisement push method in the embodiment of the present application; Figure 3 It is a schematic diagram of a physical device structure of a server in an embodiment of the present application. DETAILED DESCRIPTION

[0028] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to be used as limitations to the present application. As used in the specification of the present application, the singular expressions "one", "a kind of", "above", "the" and "this" are intended to also include plural expressions, unless there is a clear indication to the contrary in the context. It should also be understood that the term "and / or" used in the present application refers to any or all possible combinations comprising one or more of the listed items.

[0029] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise specified, "plurality" means two or more.

[0030] The following is a description of the process of the method provided by this implementation. Figure 1 , which is a flow chart of the advertising push method in an embodiment of the present application.

[0031] S101, determining a plurality of content elements in an advertisement and respective proportions of the plurality of content elements based on a preset advertisement database; The preset advertisement database refers to a data set storing multiple advertisements. Advertisement refers to an information carrier used to promote products, services or concepts. Content elements refer to the various components that make up an advertisement, such as products, spokespersons, animations, etc. The proportion is used to indicate the proportion of a content element in the entire advertisement, usually expressed as a percentage.

[0032] Specifically, first, the server accesses the preset advertisement database and obtains information about all advertisements in the preset advertisement database. Then, the server analyzes each advertisement and identifies the different content elements contained in the advertisement, such as title, text, pictures, videos, etc. Next, the server calculates the proportion (ratio) of each content element in the advertisement, which may be determined based on factors such as pixel ratio, duration ratio, or importance weight, which are not limited here. Finally, the server stores the identified content elements and their corresponding ratios for subsequent use.

[0033] The following is a list of a preset advertisement database, as well as multiple content elements in the advertisement and the corresponding proportions of the multiple content elements: Advertisement 1 Advertisement 2 Advertisement 3 Advertisement 4 Advertisement 5 …… A：70％ A：90％ B：80％ Z：60％ B：34％ B：20％ F：5％ E：17％ X：30％ X：38％ C：8％ G：3％ G：3％ Y：10％ P：28％ D：2％ H：2％ Table 1 S102, obtaining the duration of a user's stay on multiple advertisements; Here, user refers to an individual who browses or views an ad. Dwell time refers to the amount of time a user spends on an ad. Multiple ads refers to a series of different ads that a user is exposed to.

[0034] Specifically, the server performs this step in real time when the user browses the web, uses an application, or watches a video. The server captures the user's interaction information with each advertisement through front-end embedding, logging, or real-time data transmission. For static ads, the server may record the display time of the ad in the user's viewport as the length of time the user stays on the ad; for video ads, the server may record the actual length of time the user watches as the length of time the user stays on the ad. The server can also ensure that the time recorded is the time the user actually pays attention to the ad by determining the user's level of activity, such as mouse movement, clicks, or scrolling. Finally, the server summarizes and stores the length of time corresponding to each ad in preparation for subsequent analysis.

[0035] S103, determining the advertisement whose retention time exceeds the preset time as an advertisement of possible interest; The preset duration refers to a pre-set time threshold used to determine whether a user is interested in an advertisement. A potentially interesting advertisement refers to an advertisement in which a user stays for longer than the preset duration.

[0036] Specifically, the server performs this step immediately after collecting the length of time that the user stays on multiple advertisements. First, the server retrieves the preset duration, which may be a fixed value, such as 30 seconds, or it may vary depending on different types of advertisements, ad placements, or historical data analysis. The server then compares the length of time the user stays on each advertisement with the preset duration. If the user stays on an advertisement for longer than the preset duration, the server marks the advertisement as an advertisement of possible interest. This process may also involve some additional screening conditions, such as considering the user's interactive behavior (such as clicks, sharing) or the complete viewing rate of the advertisement. Finally, the server generates a list of advertisements of possible interest, which includes multiple advertisements of possible interest.

[0037] Continuing with the example in step S101, assuming that the duration of Advertisement 1 is 40 seconds, the duration of Advertisement 2 is 31 seconds, the duration of Advertisement 3 is 35 seconds, the duration of Advertisement 4 is 60 seconds, the duration of Advertisement 5 is 60 seconds, and the preset duration is 30 seconds, then Advertisement 1, Advertisement 2, Advertisement 3 and Advertisement 4 are advertisements of possible interest.

[0038] S104, determining the content element of interest to the user according to the possible content element of interest in each of the possible advertisements of interest and the corresponding proportions of the possible content element of interest; Among them, the potentially interesting content element is used to represent the various components that constitute the potentially interesting advertisement. The proportion refers to the proportion of a potentially interesting content element in the potentially interesting advertisement. The interesting content element refers to a component of the advertisement that the user is actually interested in.

[0039] Specifically, the server performs this step immediately after completing the screening of potentially interesting advertisements. First, the server extracts the potentially interesting content elements and their corresponding proportions in each potentially interesting advertisement. Then, the server may adopt a weighted calculation method, taking the user's stay time on the potentially interesting advertisement as a weight, and multiplying it by the proportion of each potentially interesting content element to obtain a weighted score for each potentially interesting content element. Next, the server may summarize the weighted scores of the same potentially interesting content elements in all potentially interesting advertisements. The server may also set a weighted score threshold, and determine the potentially interesting content elements that exceed the weighted score threshold as the user's interested content elements. Finally, the server generates a list of interested content elements, wherein the list of interested content elements includes multiple interested content elements.

[0040] Optionally, in general, determining the user's content elements of interest based on the possible content elements of interest in each of the possible advertisements of interest and the corresponding ratios of the possible content elements of interest can be achieved in the following manner: calculating the element ratio of each of the possible content elements of interest in a set of possible advertisements of interest based on the possible content elements of interest in each of the possible advertisements of interest and the corresponding ratios of the possible content elements of interest, the set of possible advertisements of interest including all possible advertisements of interest; determining the possible content elements of interest whose element ratio exceeds a preset element ratio threshold as the user's content elements of interest.

[0041] S105: Determine that the advertisement whose total proportion of the content elements of interest exceeds a preset total proportion threshold is a recommended advertisement, and push the recommended advertisement to the user.

[0042] The total proportion of the content elements of interest refers to the sum of the proportions of all the content elements of interest to the user in an advertisement. The preset total proportion threshold refers to a pre-set percentage value used to determine whether an advertisement sufficiently meets the user's interest. Recommended advertisements are used to indicate the most suitable advertisements screened by the server for display to the user. Push refers to the act of sending or displaying recommended advertisements to the user.

[0043] Specifically, the server performs this step immediately after determining the user's content elements of interest. First, the server traverses all advertisements in the preset advertisement database and calculates the total proportion of the content elements of interest in each advertisement. This process may involve weighted calculation of the importance of different content elements. Then, the server compares the calculated total proportion of the content elements of interest in each advertisement with the preset total proportion threshold. This preset total proportion threshold may vary according to different advertisement types, user groups or promotion goals. For advertisements whose total proportion of content elements of interest exceeds the preset total proportion threshold, the server marks them as recommended advertisements. Next, the server may sort these recommended advertisements, and the factors considered may include the total proportion of content elements of interest, the overall quality score of the advertisement, the bid of the advertiser, etc. Finally, the server pushes the recommended advertisements to the user based on the sorting results. The push method may include displaying on the web page browsed by the user, pushing notifications through the APP, inserting advertisements before the video is played, etc. The server may also record the user's response to the recommended advertisement, such as clicks, viewing completeness, etc., for subsequent effect analysis.

[0044] Optionally, under normal circumstances, determining that an advertisement whose total proportion of the content elements of interest exceeds a preset total proportion threshold is a recommended advertisement, and pushing the recommended advertisement to the user can be achieved in the following manner: calculating the total proportion of the content elements of interest in each advertisement in the preset advertisement database; determining the advertisement whose total proportion of the content elements of interest is higher than the preset ratio as the recommended advertisement; and pushing the recommended advertisements to the user in order of the total proportion of the content elements of interest.

[0045] By adopting the above technical solution, the server determines the user's potentially interesting advertisements based on the user's stay time on multiple advertisements. Since each potentially interesting advertisement includes multiple potentially interesting content elements and the corresponding proportions of multiple potentially interesting content elements, the server can determine the user's real interesting content elements by analyzing the user's potentially interesting advertisements. Based on these interesting content elements, the user's interesting advertisements are determined more accurately, and advertisements whose total proportion of interesting content elements exceeds the preset total proportion threshold are used as recommended advertisements and pushed to the user, achieving more accurate advertisement push, thereby increasing the probability of users clicking on advertisements, increasing the conversion rate of advertisements, and avoiding the waste of advertising resources. At the same time, this method also reduces the user's boredom with irrelevant advertisements and improves the user's experience in browsing advertisements.

[0046] The following is a more detailed description of the process of the method provided by this implementation. Figure 2 , is another flow chart of the advertising push method in the embodiment of the present application.

[0047] S201, determining a plurality of content elements in an advertisement and respective proportions of the plurality of content elements based on a preset advertisement database; For details, please refer to step S101, which will not be described in detail here.

[0048] S202, display verification information; Specifically, the server performs this step when preparing to play an advertisement, during the advertisement playing process, or at a specific time point after the advertisement playing ends. First, the server randomly selects a verification information from the preset verification information library, and the verification information is used to confirm whether the user is currently watching the advertisement. The verification information may be a simple click confirmation, a simple question and answer question, a graphic verification code, etc., or it may be a question related to the currently playing advertisement. The server then transmits the selected verification information to the user device through the network, and the verification information may be displayed on the user device in the form of a pop-up window, an overlay, or an embedded interface. The server may record the display time of the verification information in order to subsequently calculate the user response time. In addition, the server may also set the display duration of the verification information. If the user does not respond after this display duration, it may be regarded as a verification failure.

[0049] S203, receiving the user's operation response to the verification information; Specifically, the server immediately enters a waiting state after displaying the verification information. The user may respond to the verification information by clicking a button, entering text, selecting an option, or completing a specific gesture, and then the user device will transmit these operation responses to the server in real time. After receiving the operation response, the server will record the content and timestamp of the operation response. If the user's operation response is not received within a certain period of time, the server may regard it as a verification failure. The server may perform preliminary format verification and security checks on the received operation response to prevent malicious input.

[0050] S204: If the operation response is correct, continue playing the advertisement; Specifically, the server performs this step immediately after receiving the user's operation response. The server will compare the received operation response with the preset correct answer. The comparison process may involve string matching, option checking, or more complex logical judgment. If the comparison result is correct, the server will send an instruction to the user device to continue playing the advertisement. For video ads, this may mean continuing to play a paused video ad; for image ads, this may mean allowing the image ad to continue to be displayed for a specified period of time.

[0051] S205: If the operation response is wrong, stop playing the advertisement; Specifically, the server executes this step immediately after receiving the user's operation response. The server will compare the received operation response with the preset correct answer. The comparison process may involve string matching, option checking, or more complex logical judgment. If the comparison result shows an error, the server will send an instruction to the user device to stop playing the advertisement. For video ads, this means immediately terminating the playback of the video ad; for image ads, it may be removing the image ad or jumping to other pages.

[0052] S206, obtaining the duration of the user's stay on the multiple advertisements; For details, please refer to step S102, which will not be described in detail here.

[0053] S207, determining the advertisement whose retention time exceeds the preset time as an advertisement of possible interest; For details, please refer to step S103, which will not be described in detail here.

[0054] S208, determining the content element of interest to the user according to the possible content element of interest in each of the possible advertisements of interest and the corresponding proportions of the possible content element of interest; For details, please refer to step S104, which will not be described in detail here.

[0055] S209, determining that the advertisement whose total proportion of the content elements of interest exceeds a preset total proportion threshold is a recommended advertisement, and pushing the recommended advertisement to the user; For details, please refer to step S105, which will not be described in detail here.

[0056] S210, obtaining the facial expression and eye gaze duration of the user when viewing the recommended advertisement; Facial expressions refer to the changes in the user's facial muscles when watching the recommended ads, which are used to reflect the user's emotional state. Eye fixation duration refers to the duration that the user's eyes stay on a certain area of ​​the recommended ads.

[0057] Specifically, the server starts this step when the user starts watching the recommended ad, and continues to execute it throughout the entire recommended ad playback process. The server captures the user's facial expressions and eye gaze duration in real time through the front camera of the user's device. The server uses a facial recognition algorithm to analyze the user's facial expressions, which may include emotional indicators such as smiles, frowns, and surprise. The server may sample and analyze multiple times per second to capture instantaneous changes in expression. The server uses eye tracking technology to record the time the user's eyes gaze at different areas in the recommended ad, that is, the eye gaze duration. This may involve dividing the recommended ad screen into multiple areas and accurately calculating the eye gaze duration of each area. The server will store and preliminarily process the user's facial expressions and eye gaze duration when watching the recommended ad in real time in preparation for subsequent analysis. At the same time, the server may consider the user's privacy protection to ensure the anonymization and secure storage of data.

[0058] S211, determining the user's satisfaction with the recommended advertisement according to the facial expression and the eye gaze duration; Specifically, the server performs this step after obtaining enough facial expressions and eye gaze duration, which may be at the end of the advertisement or at a specific time point during the advertisement. First, the server will classify and quantify the collected facial expressions, and may assign different weight scores to different facial expressions. For example, a smile may get a higher positive score, while a frown may get a negative score. Then, the server analyzes the eye gaze duration and calculates the proportion of the user's eye gaze duration in different areas of the recommended advertisement (such as product pictures, brand logos, etc.). The server may combine these two types of data and use a preset satisfaction algorithm model to calculate a comprehensive satisfaction. This preset satisfaction algorithm model may take into account factors such as the frequency and intensity of facial expression changes and the distribution pattern of eye gaze duration.

[0059] S212: when the satisfaction score is higher than a preset score threshold, increase the push frequency of the recommended advertisement; The satisfaction score is used to represent the specific value of the user's satisfaction with the recommended advertisement. The preset score threshold represents a pre-set satisfaction score standard. The push frequency refers to the number of times or ratio of recommended advertisements displayed to users.

[0060] Specifically, the server performs this step immediately after determining the user's satisfaction score for the recommended ad. First, the server compares the calculated satisfaction score with a preset score threshold, which may vary according to different ad types, target audiences, or marketing goals. If the satisfaction score exceeds the preset score threshold, the server triggers a mechanism to increase the push frequency of the recommended ad. Specific ways of increasing the frequency may include: increasing the weight of the recommended ad in the user's ad queue, increasing the number of times the recommended ad is displayed within a specific time period, or placing the recommended ad in more display locations. The server may dynamically adjust the increase based on the extent to which the satisfaction score exceeds the preset score threshold. In addition, the server may also set an upper limit to prevent excessive push of a single ad from causing user fatigue.

[0061] S213: when the satisfaction score is lower than a preset score threshold, obtaining the reason for the user's dissatisfaction; and adjusting the advertising push strategy based on the reason for the dissatisfaction; The dissatisfaction reasons refer to the specific factors that lead users to be dissatisfied with the recommended ads. The ad push strategy is used to indicate a series of rules and methods for which ads to display to users, when to display them, and how to display them.

[0062] Specifically, after the server determines the user's satisfaction score for the recommended advertisement, it immediately executes this step. First, the server compares the calculated satisfaction score with the preset score threshold. If the satisfaction score is lower than the preset score threshold, the server needs to obtain the user's reason for dissatisfaction. The server can obtain the user's reason for dissatisfaction in a variety of ways: 1) pop up a short feedback questionnaire and let the user select or fill in the specific reason for dissatisfaction; 2) analyze the user's behavior data during the playback of the recommended advertisement, such as fast forward, mute, close and other operations; 3) combine the user's facial expressions and eye gaze duration collected previously to infer possible reasons for dissatisfaction. After obtaining the user's reasons for dissatisfaction, the server will classify and analyze the user's reasons for dissatisfaction, for example, classify the reasons for dissatisfaction into categories such as irrelevant content, too high display frequency, and poor creative quality. Then, the server adjusts the advertising push strategy based on the above analysis results.

[0063] Adjustments to advertising strategies may include: (1) If the reason for dissatisfaction is irrelevant content, the server may re-evaluate user interests and adjust the parameters of the ad recommendation algorithm.

[0064] (2) If the reason for dissatisfaction is that the display frequency is too high, the server may reduce the push frequency of this type of advertisement and increase the diversity of advertisement types.

[0065] (3) If the dissatisfaction is due to problems with the quality of the creative, the server may lower the overall weight of the ad and may also provide feedback to the advertiser, suggesting optimization of the ad creative.

[0066] S214, obtaining characteristic information of the user, the characteristic information including gender, age and dress; Specifically, the server performs this step when starting the personalized advertising recommendation process. First, the server may obtain the user's feature information through multiple channels. Feature information refers to a data set that can describe and distinguish the individual characteristics of the user. The server can: 1) retrieve the basic information provided by the user when registering, such as gender and age; 2) use computer vision technology to analyze the user's real-time image or video stream to identify gender, estimate age and analyze dressing; 3) combine the user's historical behavior data, such as shopping records, browsing history, etc., to infer the user's gender and age. For dressing, the server may use a deep learning model to identify the user's clothing style, color matching, accessories selection and other details, and then the server will standardize this information, such as converting age into age group classification and classifying dressing style into several predefined clothing types. Throughout the process, the server needs to ensure the protection of user privacy and may use data desensitization, anonymization and other technologies. In addition, the server may set up an information update mechanism to regularly refresh the user's feature information to adapt to possible changes in the user.

[0067] S215, predicting the personalized needs of the user according to the gender, age, and clothing; Specifically, the server performs this step immediately after obtaining and processing the user's feature information. The server may use a machine learning model, such as a decision tree, random forest, or neural network, to predict the user's personalized needs. Personalized needs represent specific needs or interests that meet the user's personal characteristics and preferences. This machine learning model may have been trained with a large amount of historical data and can output the category of possible personalized needs or the probability distribution of personalized needs based on the input feature information.

[0068] For example, for a middle-aged man who dresses in a business casual style, the server may predict that he has high demands in terms of career development, family life, and health and wellness.

[0069] The server may also consider environmental factors such as the current time, season, and geographic location to further adjust the predicted user's personalized needs. In addition, the server may compare the predicted user's personalized needs with the user's historical demand records to improve the accuracy and consistency of the prediction. The prediction results may be stored in the form of label sets, weight vectors, or probability distributions to provide a basis for subsequent ad matching.

[0070] S216: Determine, in the preset advertisement database, a personalized advertisement whose relevance to the personalized demand exceeds a preset relevance threshold, and push the personalized advertisement to the user.

[0071] Specifically, the server performs this step immediately after determining the user's personalized needs. First, the server traverses all advertisements in the preset advertisement database and calculates the relevance of each advertisement to the personalized needs. Advertisements whose relevance exceeds the preset relevance threshold are determined as personalized advertisements. The server sorts the filtered personalized advertisements, which may consider multiple factors such as relevance, advertiser bids, historical click-through rates, etc., and then pushes the sorted personalized advertisements to the user.

[0072] The implementation of the present invention, due to the use of dynamic interest identification and precise advertising push methods based on real-time user behavior, can accurately capture real-time changes in user interests and respond quickly, effectively solving the lag and inaccuracy problems caused by relying solely on historical data in related technologies, thereby achieving highly personalized and real-time optimization of advertising push. Due to the use of a multi-dimensional user feedback collection and analysis mechanism, it is possible to comprehensively evaluate advertising effects and continuously optimize advertising push strategies, effectively solving the problems of single user feedback and inaccurate advertising effect evaluation in related technologies, thereby achieving self-learning and continuous improvement of advertising push. Due to the use of intelligent prediction and personalized recommendation technology based on user feature information, it is possible to provide high-quality advertising services when users first contact, effectively solving the cold start problem of new users and the poor generalization of advertising recommendations in related technologies, thereby achieving rapid, accurate and wide applicability of advertising push.

[0073] The following describes the server in the embodiment of the present invention from the perspective of hardware processing. Figure 3 , is a schematic diagram of a physical device structure of a server in an embodiment of the present application.

[0074] It should be noted that Figure 3 The structure of the server shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0075] like Figure 3As shown, the server includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 302 or the program loaded from the storage part 308 to the random access memory (RAM) 303, such as executing the method described in the above embodiment. In RAM 303, various programs and data required for system operation are also stored. CPU 301, ROM 302 and RAM 303 are connected to each other through bus 304. Input / output (I / O) interface 305 is also connected to bus 304.

[0076] The following components are connected to the I / O interface 305: an input section 306 including an audio input device, a button switch, etc.; an output section 307 including a liquid crystal display (LCD) and an audio output device, an indicator light, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. A removable medium 311, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 310 as needed so that a computer program read therefrom is installed into the storage section 308 as needed.

[0077] In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication section 309, and / or installed from a removable medium 311. When the computer program is executed by a central processing unit (CPU) 301, various functions defined in the present invention are performed.

[0078] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system, apparatus, or device.

[0079] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Each box in the flowchart or block diagram may represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box may also occur in an order different from that marked in the accompanying drawings.

[0080] Specifically, the server of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, the advertisement push method provided in the above embodiment is implemented.

[0081] As another aspect, the present invention further provides a computer-readable storage medium, which may be included in the server described in the above embodiment; or may exist independently without being assembled into the server. The above storage medium carries one or more computer programs, and when the above one or more computer programs are executed by a processor of the server, the server implements the advertisement push method provided in the above embodiment.

[0082] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0083] As used in the above embodiments, the term "when..." may be interpreted to mean "if..." or "after..." or "in response to determining..." or "in response to detecting...", depending on the context. Similarly, the phrases "upon determining..." or "if (the stated condition or event) is detected" may be interpreted to mean "if determining..." or "in response to determining..." or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)", depending on the context.

[0084] Those skilled in the art can understand that to implement all or part of the processes in the above-mentioned embodiments, the processes can be completed by computer programs to instruct related hardware, and the programs can be stored in computer-readable storage media. When the programs are executed, they can include the processes of the above-mentioned method embodiments. The aforementioned storage media include: ROM or random access memory RAM, magnetic disk or optical disk and other media that can store program codes.

Claims

1. An advertisement push method, characterized in that: Applied to a server, the method comprises: Based on a preset advertisement database, determining a plurality of content elements in the advertisement and respective proportions of the plurality of content elements; Obtaining the length of time a user spends on the plurality of advertisements; Determine the advertisement whose retention time exceeds the preset time as an advertisement of possible interest; Determine the content element of interest to the user according to the possible content element of interest in each of the possible advertisements of interest and the corresponding proportions of the possible content element of interest; Determine an advertisement whose total proportion of the interested content elements exceeds a preset total proportion threshold as a recommended advertisement, and push the recommended advertisement to the user.

2. The method according to claim 1, characterized in that The determining of the user's content elements of interest based on the content elements of possible interest in each of the advertisements of possible interest and the corresponding proportions of the content elements of possible interest specifically includes: Based on the possibly interesting content elements in each of the possibly interesting advertisements and the corresponding proportions of the possibly interesting content elements, calculating the proportion of each of the possibly interesting content elements in the possibly interesting advertisement set, the possibly interesting advertisement set including all the possibly interesting advertisements; The possible content elements of interest whose element proportion exceeds a preset element proportion threshold are determined as the content elements of interest to the user.

3. The method according to claim 1, characterized in that The determining that the advertisement whose total proportion of the content elements of interest exceeds a preset total proportion threshold is a recommended advertisement, and pushing the recommended advertisement to the user specifically includes: Calculating the total proportion of the content elements of interest in each advertisement in the preset advertisement database; Determine the advertisement whose total proportion of the content elements of interest is higher than the preset ratio as the recommended advertisement; The recommended advertisements are pushed to the user in order according to the total proportion of the interested content elements.

4. The method according to claim 1, characterized in that: Before the step of obtaining the duration of the user's stay on the plurality of advertisements, the method further includes: Display verification information; receiving an operation response of the user to the verification information; If the operation response is correct, continue playing the advertisement; If the operation response is an error, the advertisement playing is stopped.

5. The method according to claim 1, characterized in that After the step of determining that the advertisement whose total proportion of the content elements of interest exceeds a preset total proportion threshold is a recommended advertisement, and pushing the recommended advertisement to the user, the method further includes: Acquire the facial expression and eye gaze duration of the user when watching the recommended advertisement; determining the user's satisfaction with the recommended advertisement according to the facial expression and the eye gaze duration; When the satisfaction score is higher than a preset score threshold, the push frequency of the recommended advertisement is increased.

6. The method according to claim 5, characterized in that After the step of determining the user's satisfaction with the recommended advertisement according to the facial expression and the eye gaze duration, the method further includes: When the satisfaction score is lower than a preset score threshold, obtaining the reason for the user's dissatisfaction; The advertising push strategy is adjusted based on the reasons for dissatisfaction.

7. The method according to claim 1, characterized in that The method further comprises: Acquire user's characteristic information, including gender, age and dress; Predicting the personalized needs of the user according to the gender, age and clothing; In the preset advertisement database, a personalized advertisement whose relevance to the personalized demand exceeds a preset relevance threshold is determined, and the personalized advertisement is pushed to the user.

8. A server, characterized in that: The server includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the server to execute the method described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that: When the instructions are executed on a server, the server is caused to execute the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that When the computer program product is run on a server, the server is caused to execute the method according to any one of claims 1 to 7.

Images