Advertisement loading method and system based on user behavior

Abstract

The present invention relates to the field of advertisement loading, and in particular to a method and system for advertisement loading based on user behavior. The present invention stores historical user behavior characteristics for a user terminal, determines tendency characteristic operation data for the user terminal based on the historical user behavior characteristics, determines an advertisement browsing frequency characterization coefficient according to a comparison result between the user terminal's current operation data for a browsing page and the tendency characteristic operation data, so as to determine the frequent browsing tendency category of the user terminal, and performs advertisement loading and pushing for the user terminal at the push position based on the frequent browsing tendency category of the user terminal. The present invention takes into account the differences in operation data of the user terminal before showing purchase tendency, performs personalized analysis for different user terminals, and adaptively loads and pushes advertisements in different ways, thereby improving the accuracy of advertisement pushing.

Description

Technical Field

[0001] The present invention relates to the field of advertisement loading, and in particular to a method and system for advertisement loading based on user behavior. Background Art

[0002] The popularity of the Internet has prompted more and more users to conduct activities online. For this reason, these activities have generated a large amount of behavioral data that can reflect users' purchasing tendencies, providing a large number of reference resources for advertising providers. At the same time, when users browse the Internet, they prefer to see advertisements related to the content they are interested in, rather than the traditional wide-ranging form of advertising content that is less relevant to them. This situation has promoted the development of user behavior-based advertising loading and pushing technology. Advertisers pursue higher advertising click-through rates and purchase conversion rates, which enables them to formulate more accurate advertising loading and pushing strategies by analyzing user behavior.

[0003] Chinese Patent Publication No.: CN112073525A discloses an advertisement push method, device and electronic device, which are applied to mobile terminals. The method includes: obtaining an advertisement strategy combination for storage, wherein the advertisement strategy combination includes at least two cache advertisement strategies with different configuration information; generating an advertisement slot push request based on a preload strategy of the configuration information; obtaining advertisement content of the cache advertisement strategy corresponding to the advertisement slot push request according to the advertisement slot push request; and displaying the advertisement content in a target advertisement slot. The present invention provides a combination mode of multiple advertisement strategies, wherein the configuration information between the advertisement strategies is different, and a delivery mode of multiple advertisement strategies for the same advertisement slot can meet different needs, such as using different advertisement strategies in different time periods, and advertisers can arbitrarily arrange and combine the advertisement strategies.

[0004] However, there are still the following problems in the prior art:

[0005] Due to differences in habits, different users have different operating habits for browsing pages, and thus, the related browsing operations performed before purchasing goods will also be different. Therefore, when analyzing purchase tendencies through big data, different users perform differently in data dimensions. The existing technology does not take individual differences among users into account, resulting in poor accuracy when pushing advertisements. Summary of the invention

[0006] To this end, the present invention provides an advertisement loading method and system based on user behavior, so as to overcome the problem that when purchasing tendency is analyzed through big data, different user terminals have different performances in data dimensions, and the prior art does not take individual differences of user terminals into consideration, resulting in poor accuracy when pushing advertisements.

[0007] To achieve the above object, the present invention provides an advertisement loading method based on user behavior, which comprises:

[0008] Storing historical user behavior characteristics for a user terminal, the historical user behavior characteristics including operation data for browsing pages within a predetermined time before the user terminal completes a purchase operation;

[0009] Determining characteristic operation data for the user terminal based on the historical user behavior characteristics;

[0010] Determine the advertisement browsing frequency characterization coefficient according to the comparison result of the operation data of the user terminal currently browsing the page and the tendency characteristic operation data, so as to determine the browsing frequency tendency category of the user terminal;

[0011] Determine a push location for the advertisement based on the historical behavior characteristics;

[0012] Based on the frequent browsing tendency category of the user terminal, the advertisement is loaded and pushed to the user terminal at the push location, including:

[0013] Determine the key triggering event of the user terminal, extract the browsing coverage of the user terminal for the detail page corresponding to the advertisement after the key triggering event, and / or obtain the viewing time of the evaluation content of the advertisement by the user terminal to calculate the purchase intention representation value, so as to determine whether the purchase intention stage has been reached, determine the operation data of the user terminal for similar advertisements, analyze the search depth, and adjust the loading and pushing frequency of the advertisement accordingly;

[0014] Or, obtain several advertisements, randomly select them, and then load and push them;

[0015] The key triggering event is that the user clicks on the detail page of the advertisement and views the evaluation content of the advertisement.

[0016] Furthermore, the process of determining the tendency characteristic operation data for the user terminal based on the historical user behavior characteristics includes:

[0017] Call historical user behavior characteristics to obtain certain operation data, including the number of times the user browses the advertisements associated with the purchased products and the browsing time;

[0018] Calculate the average number of views and average browsing time for each purchased product;

[0019] The average browsing times and the average browsing time are set as the tendency characteristic operation data.

[0020] Furthermore, the process of determining the advertisement browsing frequency characterization coefficient according to the comparison result of the user terminal's current operation data on the browsing page and the tendency characteristic operation data includes:

[0021] Record the number of times the user has viewed the advertisement and the duration of the current viewing;

[0022] Determine the ratio of the current browsing times to the average browsing times as the first browsing feature;

[0023] Determine the ratio of the current browsing time to the average browsing time as the second browsing feature;

[0024] The sum of the first browsing feature and the second browsing feature is determined as the advertisement browsing frequency characterization coefficient.

[0025] Further, determining the frequent browsing tendency category of the user terminal includes:

[0026] If the advertisement browsing frequency representation coefficient is greater than or equal to the advertisement browsing frequency representation coefficient threshold, the user terminal is determined to be a high frequency tendency category;

[0027] If the advertisement browsing frequency representation coefficient is less than the advertisement browsing frequency representation coefficient threshold, it is determined that the user terminal is a low frequency tendency category.

[0028] Furthermore, the process of determining the push location for the advertisement based on the historical behavior characteristics includes:

[0029] Call historical user behavior characteristics to obtain certain operation data, including the location of the clicked ad on the browsing page;

[0030] Determine the probability of the clicked advertisement appearing in each area of ​​the browsing page based on the operation data;

[0031] The area corresponding to the maximum occurrence probability is determined as the advertisement push location.

[0032] Further, loading and pushing advertisements to the user terminal at the push location includes:

[0033] If the user terminal is a high-frequency tendency category, determine the key triggering event of the user terminal to extract the browsing coverage of the user terminal for the detail page corresponding to the advertisement after the key triggering event or / and obtain the viewing time of the evaluation content of the advertisement by the user terminal to calculate the purchase intention representation value to determine whether the purchase intention stage has been reached, determine the operation data of the user terminal for similar advertisements, analyze the search depth, and adjust the loading and pushing frequency of the advertisement accordingly;

[0034] If the user terminal is of the low-frequency tendency category, a number of advertisements are obtained for random screening and advertisement loading and push.

[0035] Furthermore, the process of calculating the purchase intention representation value includes:

[0036] If the key triggering event of the user terminal is that the user terminal clicks on the detail page of the advertisement, extracting the browsing coverage of the detail page of the advertisement by the user terminal, wherein the ratio of the browsing coverage of the detail page of the advertisement to the browsing coverage threshold is determined as the purchase intention representation value;

[0037] If the key triggering event of the user terminal is viewing the evaluation content of the advertisement, then extract the browsing coverage of the user terminal for the detail page of the advertisement and obtain the viewing time of the evaluation content of the advertisement,

[0038] Determine the ratio of the page-reading coverage of the detail page to the page-reading coverage threshold as the first intention feature,

[0039] Determine the ratio of the viewing time of the evaluation content to the viewing time threshold as the second intention feature,

[0040] The sum of the first intention feature and the second intention feature is determined as the purchase intention representation value.

[0041] Further, determining whether the purchase intention stage has been reached includes:

[0042] If the purchase intention representation value is greater than or equal to the purchase intention representation threshold, it is determined that the purchase intention stage has been reached.

[0043] Furthermore, the process of determining the user's operation data for similar advertisements and analyzing the search depth includes:

[0044] Obtaining the user's operation data on similar advertisements after reaching the purchase intention stage, including the continuous browsing time of the user on similar advertisements;

[0045] The ratio of the continuous browsing time to a benchmark continuous browsing time is determined as the search depth.

[0046] Further, adjusting the loading and pushing frequency of the advertisement includes:

[0047] The loading and pushing frequency of advertisements is increased, and the increase in the loading and pushing frequency is positively correlated with the search depth.

[0048] Compared with the prior art, the present invention stores historical user behavior characteristics for a user terminal, determines tendency characteristic operation data for the user terminal based on the historical user behavior characteristics, determines the advertisement browsing frequency characterization coefficient according to the comparison result between the user terminal's current operation data for the browsing page and the tendency characteristic operation data, so as to determine the frequent browsing tendency category of the user terminal, and loads and pushes advertisements for the user terminal at the push location based on the frequent browsing tendency category of the user terminal. The present invention takes into account the differences in operation data of the user terminal before showing purchase tendency, conducts personalized analysis for different user terminals, and adaptively loads and pushes advertisements in different ways, thereby improving the accuracy of advertisement push.

[0049] In particular, the present invention stores the operation data for browsing pages within a predetermined time before the user terminal completes the purchase operation, and determines the tendency characteristic operation data. In actual situations, there are habit differences between different user terminals, and thus, before completing the purchase operation, the related operations performed by the user terminal are different. Therefore, the present invention analyzes historical user behavior characteristics to determine the tendency characteristic operation data unique to the user terminal for subsequent determination of the advertisement browsing frequency characterization coefficient, and thus can more accurately identify the time nodes when the user terminal has a higher intention, and subsequently adaptively adjust the advertisement loading and push method to improve the accuracy and efficiency of advertisement push.

[0050] In particular, the present invention considers determining the push location for advertisements through historical behavior characteristics, and considering the habit differences between user terminals, the personalized user terminal uses the area that is easy to click before completing the purchase operation as the advertisement push location, thereby improving the advertisement push effect.

[0051] In particular, the present invention determines the key trigger events for the user terminals of the high-frequency tendency category, calculates the purchase intention representation value based on the key trigger events to represent the user's interest in the product and purchase intention, and then determines whether it has reached the intention to purchase stage; in actual situations, the user terminal will make a comprehensive consideration by viewing the details interface and evaluation content of the advertisement to further make a decision on whether there is a purchase tendency. Therefore, the key trigger events involved in the above situation are viewing the details page and viewing the evaluation content of the advertisement. By analyzing whether the user has a purchase intention through more representative behavioral characteristics, the relevance of the loaded and pushed advertisements is guaranteed, thereby improving the advertisement loading and push effect.

[0052] In particular, the present invention determines the search depth for situations where the intention to purchase stage is reached. In actual situations, the user side has a multi-dimensional comparison of commodities with a strong purchasing tendency. Therefore, the above situation is characterized by the search depth to further determine the purchasing tendency of the user side, and adaptively adjust the frequency of advertisement loading and pushing, to ensure the accuracy of advertisement loading and pushing, and to improve the effect of advertisement loading and pushing. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A schematic diagram of the steps of a method for loading advertisements based on user behavior according to an embodiment of the invention;

[0054] Figure 2 A logical decision diagram for determining the frequent browsing tendency category of a user terminal for an embodiment of the invention;

[0055] Figure 3 A logical decision diagram for loading and pushing advertisements at a push position in an embodiment of the invention;

[0056] Figure 4 A logical decision diagram for determining whether the purchase intention stage has been reached according to an embodiment of the present invention. DETAILED DESCRIPTION

[0057] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0058] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0059] See also Figures 1 to 4 As shown, Figure 1 is a schematic diagram of steps of a method for loading advertisements based on user behavior according to an embodiment of the present invention, Figure 2 A logical decision diagram for determining the frequent browsing tendency category of a user terminal in an embodiment of the present invention, Figure 3 This is a logic decision diagram for loading and pushing advertisements at a push position according to an embodiment of the present invention. Figure 4 The following is a logic decision diagram for determining whether the purchase intention stage has been reached in an embodiment of the present invention. The method for loading advertisements based on user behavior in an embodiment of the present invention includes:

[0060] S1, storing historical user behavior characteristics for a user terminal, wherein the historical user behavior characteristics include operation data for browsing pages within a predetermined period before the user terminal completes a purchase operation;

[0061] S2, determining the tendency characteristic operation data for the user terminal based on the historical user behavior characteristics;

[0062] S3, determining an advertisement browsing frequency characterization coefficient according to a comparison result between the user terminal's current operation data on the browsing page and the tendency characteristic operation data, so as to determine the user terminal's browsing frequency tendency category;

[0063] S4, determining a push location for the advertisement based on the historical behavior characteristics;

[0064] S5, based on the frequent browsing tendency category of the user terminal, loading and pushing advertisements to the user terminal at the push location, including:

[0065] Determine the key triggering event of the user terminal, extract the browsing coverage of the user terminal for the detail page corresponding to the advertisement after the key triggering event, and / or obtain the viewing time of the evaluation content of the advertisement by the user terminal to calculate the purchase intention representation value, so as to determine whether the purchase intention stage has been reached, determine the operation data of the user terminal for similar advertisements, analyze the search depth, and adjust the loading and pushing frequency of the advertisement accordingly;

[0066] Or, obtain a number of advertisements, randomly screen them, load and push them, and load and push the randomly screened advertisements;

[0067] The key triggering event is that the user clicks on the detail page of the advertisement and views the evaluation content of the advertisement.

[0068] It is understandable that there is no limitation on the method of storing the historical user behavior characteristics of the user terminal. The relevant information is collected and stored only after obtaining the authorization of the user terminal, which will not be elaborated here.

[0069] Specifically, there is no limitation on the specific way of browsing the page. It can be any interface displayed by any terminal, such as a shopping mall page, a short video push page, etc., which will not be repeated here.

[0070] Specifically, in order to characterize the validity of the data, the predetermined time is set within the interval of [30min, 60min].

[0071] Specifically, the process of determining the tendency characteristic operation data for the user terminal based on the historical user behavior characteristics includes:

[0072] Call historical user behavior characteristics to obtain certain operation data, including the number of times the user browses the advertisements associated with the purchased products and the browsing time;

[0073] Calculate the average number of views and average browsing time for each purchased product;

[0074] The average browsing times and the average browsing time are set as the tendency characteristic operation data.

[0075] It is understandable that completing a single purchase corresponds to a single purchase data, and the advertisements associated with the purchased product are advertisements corresponding to other products of the same type as the purchased product.

[0076] Taking the purchase of model A TV as an example, the advertisements associated with the purchase of the product are advertisements corresponding to other models of TVs, which will not be repeated here.

[0077] Specifically, the scrolling coverage is the ratio of the area of ​​the detail page viewed by the user to the total area of ​​the detail page.

[0078] The present invention stores the operation data for browsing pages within a predetermined time before the user terminal completes the purchase operation, and determines the tendency characteristic operation data. In actual situations, there are habit differences between different user terminals, and thus, before completing the purchase operation, the related operations performed by the user terminal are different. Therefore, the present invention analyzes the historical user behavior characteristics, determines the tendency characteristic operation data unique to the user terminal, and uses it to subsequently determine the advertisement browsing frequency characterization coefficient, and thus can more accurately identify the time nodes when the user terminal has a higher intention, and subsequently adaptively adjust the advertisement loading and pushing method to improve the accuracy and efficiency of advertisement pushing.

[0079] Specifically, the process of determining the advertisement browsing frequency representation coefficient based on the comparison result of the user's current operation data on the browsing page and the tendency characteristic operation data includes:

[0080] Record the number of times the user has viewed the advertisement and the duration of the current viewing;

[0081] Determine the ratio of the current browsing times to the average browsing times as the first browsing feature;

[0082] Determine the ratio of the current browsing time to the average browsing time as the second browsing feature;

[0083] The sum of the first browsing feature and the second browsing feature is determined as the advertisement browsing frequency characterization coefficient.

[0084] Specifically, determining the frequent browsing tendency category of the user terminal includes:

[0085] If the advertisement browsing frequency representation coefficient is greater than or equal to the advertisement browsing frequency representation coefficient threshold, the user terminal is determined to be a high frequency tendency category;

[0086] If the advertisement browsing frequency representation coefficient is less than the advertisement browsing frequency representation coefficient threshold, it is determined that the user terminal is a low frequency tendency category.

[0087] The threshold of the advertisement browsing frequency characterization coefficient is selected in the interval [2.05, 2.15].

[0088] Specifically, the process of determining the push location for the advertisement based on the historical behavior characteristics includes:

[0089] Call historical user behavior characteristics to obtain certain operation data, including the location of the clicked ad on the browsing page;

[0090] Determine the probability of the clicked advertisement appearing in each area of ​​the browsing page based on the operation data;

[0091] The area corresponding to the maximum occurrence probability is determined as the advertisement push location.

[0092] It can be understood that the appearance probability of a single area is the ratio of the number of clicked advertisements in the area to the total number of clicked advertisements.

[0093] The present invention considers determining the push location for advertisements through historical behavior characteristics, and considering the habit differences between user terminals, a personalized user terminal uses an area that is easy to click before completing a purchase operation as an advertisement push location, thereby improving the advertisement push effect.

[0094] Specifically, loading and pushing advertisements to the user terminal at the push location includes:

[0095] If the user terminal is a high-frequency tendency category, determine the key triggering event of the user terminal to extract the browsing coverage of the user terminal for the detail page corresponding to the advertisement after the key triggering event or / and obtain the viewing time of the evaluation content of the advertisement by the user terminal to calculate the purchase intention representation value to determine whether the purchase intention stage has been reached, determine the operation data of the user terminal for similar advertisements, analyze the search depth, and adjust the loading and pushing frequency of the advertisement accordingly;

[0096] If the user terminal is of the low-frequency tendency category, a number of advertisements are obtained for random screening and advertisement loading and push.

[0097] Specifically, the process of calculating the purchase intention representation value includes:

[0098] If the key triggering event of the user terminal is that the user terminal clicks on the detail page of the advertisement, extracting the browsing coverage of the detail page of the advertisement by the user terminal, wherein the ratio of the browsing coverage of the detail page of the advertisement to the browsing coverage threshold is determined as the purchase intention representation value;

[0099] If the key triggering event of the user terminal is viewing the evaluation content of the advertisement, then extract the browsing coverage of the user terminal for the detail page of the advertisement and obtain the viewing time of the evaluation content of the advertisement,

[0100] Determine the ratio of the page-reading coverage of the detail page to the page-reading coverage threshold as the first intention feature,

[0101] Determine the ratio of the viewing time of the evaluation content to the viewing time threshold as the second intention feature,

[0102] The sum of the first intention feature and the second intention feature is determined as the purchase intention representation value.

[0103] Specifically, the browsing coverage threshold of the details page and the viewing time threshold of the evaluation content are pre-set, and relevant data of several user terminals browsing the platform are obtained, the browsing coverage data of the details page and the viewing time data of the evaluation content are called, the average browsing coverage and the average viewing time are solved, and the browsing coverage threshold is set to between 1.35 and 1.45 times the average browsing coverage, and the viewing time threshold is set to between 1.45 and 1.65 times the average viewing time.

[0104] The present invention determines the key trigger events for the user terminals of the high-frequency tendency category, calculates the purchase intention representation value based on the key trigger events to represent the user's interest in and purchase intention of the product, and then determines whether the purchase intention stage has been reached; in actual situations, the user terminal will make a comprehensive consideration by viewing the details interface of the advertisement and the evaluation content to further make a decision on whether there is a purchase tendency. Therefore, the key trigger events involved in the above situation are viewing the details page and viewing the evaluation content of the advertisement. Whether the user has a purchase intention is analyzed through more representative behavioral characteristics, and the relevance of the loaded and pushed advertisements is guaranteed, thereby improving the advertisement loading and push effect.

[0105] Specifically, determining whether the purchase intention stage has been reached includes:

[0106] If the purchase intention representation value is greater than or equal to the purchase intention representation threshold, it is determined that the purchase intention stage has been reached.

[0107] For the purchase intention representation value determined based on the ratio of the flipping coverage to the flipping coverage threshold, the purchase intention representation threshold is selected within the interval [50%, 60%] for comparison;

[0108] For the purchase intention representation value determined based on the sum of the first intention feature and the second intention feature, the purchase intention representation threshold is selected within the interval [2.35, 2.45] for comparison.

[0109] The present invention determines the search depth for the situation of reaching the purchase intention stage. In actual situations, the user side has a multi-dimensional comparison of commodities with a strong purchase tendency. Therefore, the above situation is characterized by the search depth to further determine the purchase tendency of the user side, and adaptively adjust the frequency of advertisement loading and pushing, to ensure the accuracy of advertisement loading and pushing, and improve the effect of advertisement loading and pushing.

[0110] Specifically, the process of determining the user's operational data for similar advertisements and analyzing the search depth includes:

[0111] Obtaining the user's operation data on similar advertisements after reaching the purchase intention stage, including the continuous browsing time of the user on similar advertisements;

[0112] The ratio of the continuous browsing time to a benchmark continuous browsing time is determined as the search depth.

[0113] It can be understood that the continuous browsing time is the duration of the user's continuous and uninterrupted browsing of a single type of advertisement.

[0114] The benchmark continuous browsing time is determined based on the historical user behavior characteristics of the user end.

[0115] The historical user behavior characteristics are called to obtain a number of operation data, including a number of continuous browsing durations of the user end, to solve the average of the continuous browsing durations, and to determine the average of the continuous browsing durations as the benchmark continuous browsing duration.

[0116] Specifically, adjusting the loading and pushing frequency of the advertisement includes:

[0117] The loading and pushing frequency of advertisements is increased, and the increase in the loading and pushing frequency is positively correlated with the search depth.

[0118] The loading push frequency is the number of ads pushed per unit time, where the unit time is 1 minute.

[0119] Specifically, in implementation, optionally,

[0120] If the search depth is in the interval [1, 1.15), the load push frequency increase is 0.15 times the initial load push frequency;

[0121] If the search depth is in the interval [1.15, 1.25), the load push frequency increase is 0.25 times the initial load push frequency;

[0122] If the search depth is in the interval [1.25, +∞), the load push frequency increase is 0.35 times the initial load push frequency.

[0123] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A method for loading advertisements based on user behavior, characterized in that: include: Storing historical user behavior characteristics for a user terminal, the historical user behavior characteristics including operation data for browsing pages within a predetermined time before the user terminal completes a purchase operation; Determining characteristic operation data for the user terminal based on the historical user behavior characteristics; Determine the advertisement browsing frequency characterization coefficient according to the comparison result of the operation data of the user terminal currently browsing the page and the tendency characteristic operation data, so as to determine the browsing frequency tendency category of the user terminal; Determine a push location for the advertisement based on the historical behavior characteristics; Based on the frequent browsing tendency category of the user terminal, the advertisement is loaded and pushed to the user terminal at the push location, including: Determine the key triggering event of the user terminal, extract the browsing coverage of the user terminal for the detail page corresponding to the advertisement after the key triggering event, and / or obtain the viewing time of the evaluation content of the advertisement by the user terminal to calculate the purchase intention representation value, so as to determine whether the purchase intention stage has been reached, determine the operation data of the user terminal for similar advertisements, analyze the search depth, and adjust the loading and pushing frequency of the advertisement accordingly; Or, obtain several advertisements, randomly select them, and then load and push them; The key triggering event is the user terminal clicking on the detail page of the advertisement and viewing the evaluation content of the advertisement; The process of determining the push location for the advertisement based on the historical behavior characteristics includes: Call historical user behavior characteristics to obtain certain operation data, including the location of the clicked ad on the browsing page; Determine the probability of the clicked advertisement appearing in each area of ​​the browsing page based on the operation data; The area corresponding to the maximum occurrence probability is determined as the advertisement push location; Determine the user's operational data for similar ads and analyze the search depth process includes: Obtaining the user's operation data on similar advertisements after reaching the purchase intention stage, including the continuous browsing time of the user on similar advertisements; Determine the ratio of the continuous browsing time to the benchmark continuous browsing time as the search depth; The process of determining the tendency characteristic operation data for the user terminal based on the historical user behavior characteristics includes: Call historical user behavior characteristics to obtain certain operation data, including the number of times the user browses the advertisements associated with the purchased products and the browsing time; Calculate the average number of views and average browsing time for each purchased product; The average of the number of browsing times and the average of the browsing time are set as the tendency characteristic operation data; The process of calculating the purchase intent representation value includes: If the key triggering event of the user terminal is that the user terminal clicks on the detail page of the advertisement, extracting the browsing coverage of the detail page of the advertisement by the user terminal, wherein the ratio of the browsing coverage of the detail page of the advertisement to the browsing coverage threshold is determined as the purchase intention representation value; If the key triggering event of the user terminal is viewing the evaluation content of the advertisement, then extract the browsing coverage of the user terminal for the detail page of the advertisement and obtain the viewing time of the evaluation content of the advertisement, Determine the ratio of the page-reading coverage of the detail page to the page-reading coverage threshold as the first intention feature, Determine the ratio of the viewing time of the evaluation content to the viewing time threshold as the second intention feature, The sum of the first intention feature and the second intention feature is determined as the purchase intention representation value.

2. The method for loading advertisements based on user behavior according to claim 1, characterized in that: The process of determining the advertisement browsing frequency characterization coefficient based on the comparison result of the user's current operation data on the browsing page and the tendency characteristic operation data includes: Record the number of times the user has viewed the advertisement and the duration of the current viewing; Determine the ratio of the current browsing times to the average browsing times as the first browsing feature; Determine the ratio of the current browsing time to the average browsing time as the second browsing feature; The sum of the first browsing feature and the second browsing feature is determined as the advertisement browsing frequency characterization coefficient.

3. The method for loading advertisements based on user behavior according to claim 1, characterized in that: Determining the frequent browsing tendency category of the user terminal includes: If the advertisement browsing frequency representation coefficient is greater than or equal to the advertisement browsing frequency representation coefficient threshold, the user terminal is determined to be a high frequency tendency category; If the advertisement browsing frequency representation coefficient is less than the advertisement browsing frequency representation coefficient threshold, it is determined that the user terminal is a low frequency tendency category.

4. The method for loading advertisements based on user behavior according to claim 1, characterized in that: Loading and pushing advertisements to the user terminal at the push location includes: If the user terminal is a high-frequency tendency category, determine the key triggering event of the user terminal to extract the browsing coverage of the user terminal for the detail page corresponding to the advertisement after the key triggering event or / and obtain the viewing time of the evaluation content of the advertisement by the user terminal to calculate the purchase intention representation value to determine whether the purchase intention stage has been reached, determine the operation data of the user terminal for similar advertisements, analyze the search depth, and adjust the loading and pushing frequency of the advertisement accordingly; If the user terminal is of the low-frequency tendency category, a number of advertisements are obtained for random screening and advertisement loading and push.

5. The method for loading advertisements based on user behavior according to claim 1, characterized in that: Determining whether the purchase intention stage has been reached includes: If the purchase intention representation value is greater than or equal to the purchase intention representation threshold, it is determined that the purchase intention stage has been reached.

6. The method for loading advertisements based on user behavior according to claim 1, characterized in that: Adjusting the loading push frequency for the advertisement includes: The loading and pushing frequency of advertisements is increased, and the increase in the loading and pushing frequency is positively correlated with the search depth.

Images