Game advertisement putting method based on user portrait

Abstract

The invention relates to the field of game advertisement putting, in particular to a user portrait-based game advertisement putting method, which comprises the following steps of: extracting corresponding daily input characteristics according to input data of a user side for a game; and in combination with the game theme browsing duration ratio of the user side on the vertical platform, the conversion value characterization parameter of the user side is analyzed so as to divide the corresponding value conversion potential category, and the game advertisement putting of the user side is adaptively adjusted and analyzed. According to the method, the interest of the user is accurately locked through dual mechanisms, so that the push content of the game advertisement better fits the real interest of the user side, the acceptance and experience of the user side on the advertisement are improved, invalid interference is reduced, the push content is matched with the updating rhythm of the game, the timeliness of the advertisement is synchronized with the demand, and the dynamic interest change of the user side is met. Meanwhile, the advertisement conversion efficiency is improved, the putting cost is reduced, the advertisement ecological quality of the vertical platform is improved, and the user side stickiness is enhanced.

Description

Technical Field

[0001] This invention relates to the field of game advertising, and more particularly to a method for game advertising based on user profiles. Background Technology

[0002] With the rapid development of mobile internet technology, the gaming industry has experienced explosive growth, with a surge in the number of game products and increasingly fierce market competition. Against this backdrop, advertising has become one of the core methods for game developers to increase product exposure, attract new users, and drive user conversion to paid subscriptions. The effectiveness of advertising directly impacts the market performance and commercial revenue of game products.

[0003] In the early stages of game advertising, a broad-based approach was often adopted, which involved sending uniform advertising content to a massive number of users across various platforms. This model did not require precise user segmentation, had a simple process, and low upfront costs. It played a certain role in promoting games when product availability was scarce and user demand was relatively concentrated.

[0004] As the user base continues to expand, user preferences for games are becoming increasingly diversified and personalized. Different users exhibit significant differences in game genres (role-playing, competitive, casual puzzle, etc.), gameplay, and willingness to pay. Simultaneously, vertical platforms are rapidly emerging. These platforms focus on the gaming sector, aggregating a large number of targeted game users and accumulating rich user behavior data, providing a data foundation for precise advertising.

[0005] To improve advertising efficiency and reduce resource waste, targeted advertising is gradually becoming an industry trend. User profiling, as the core support for targeted advertising, integrates various user data to build characteristic models of user interests, preferences, and behavioral habits, helping game developers better understand user needs.

[0006] Chinese Patent Application Publication No. CN117436957A discloses a method for precise advertising of game software, including: collecting the percentage of user playback time and frequency of each type of video within a reference period at each time point; obtaining the user's preference promotion rate for each type of video at each time point; plotting a preference promotion rate curve; obtaining the user's most recent preference data points for each type of video based on the differences between data points in the preference promotion rate curve; further obtaining the degree of preference change and the probability of preference shift for each type of video; obtaining the consistency between the trend of the preference promotion rate curve for each type of video and user preferences based on the probability of preference shift for each type of video; and combining the degree of preference change to obtain the preference recommendation degree for each type of video for precise advertising recommendation. This invention makes advertising recommendations more accurate and reduces the waste of advertising resources.

[0007] However, the following problems still exist in the existing technology. The common practice of using unified push notifications or targeting based on a single data dimension leads to a mismatch between ad content and user interests, reducing the accuracy of game ad targeting. Furthermore, it fails to consider that user interests are constantly evolving, making it difficult to keep up with shifts in user focus due to game updates and adapting to the dynamic needs of users. Summary of the Invention

[0008] To address this, the present invention provides a game advertising delivery method based on user profiles, which overcomes the problems of mismatch between advertising content and user interests in existing technologies that use unified push or delivery based on a single data dimension, thereby reducing the accuracy of game advertising delivery, and failing to consider that user interests are in a dynamic process, making it impossible to keep up with the shift in user attention brought about by game updates, and making it difficult to adapt to game updates and dynamic user needs.

[0009] To achieve the above objectives, the present invention provides a method for game advertising delivery based on user profiles, comprising: Based on user data on game engagement, corresponding daily engagement features are extracted. These features include the ratio of active time to online time within a predetermined period and the deviation value of the time interval between online status switching. By combining the daily input characteristics and the proportion of browsing time of game-related content on the vertical platform, the conversion value representation parameters of the user terminal are analyzed to classify the value conversion potential categories of the user terminal. Based on the aforementioned value conversion potential categories, adjustments and analyses are made to the game advertising placement on the user's end, including: Identify the associated user terminals of the user terminal, determine the focus of attention between the user terminal and the associated user terminals, determine the fit between the update frequency and the change frequency of the focus of attention based on the update frequency of the game and the change frequency, and determine the update attention characterization value of the user terminal for the game by combining the time difference of the interaction behavior corresponding to the focus of attention. In order to adjust the game advertising delivery method for the user terminal. Game ads are randomly displayed on vertical platforms at unit frequencies.

[0010] Furthermore, the process of analyzing the conversion value representation parameters of the user end includes: The sum of the ratio of active time to online time within a predetermined period to the time percentage threshold and the ratio of the time interval deviation threshold to the time interval deviation value of online status switching is used as the first conversion value feature. The ratio of the percentage of time users spend browsing game-related content on vertical platforms to the threshold percentage of time users spend browsing game-related content is used as the second conversion value feature. The first conversion value feature and the second conversion value feature are weighted and summed to form the conversion value representation parameter.

[0011] Furthermore, the value conversion potential categories of the aforementioned user terminals are divided into: If the conversion value representation parameter of the user terminal is greater than or equal to the conversion value representation parameter threshold, then the user terminal is classified as a high-potential value conversion category. If the conversion value representation parameter of the user terminal is less than the conversion value representation parameter threshold, then the user terminal is classified as a low-potential value conversion category.

[0012] Furthermore, the analysis and adjustment of game advertising delivery on the user's end includes: If the user terminal is a category with high potential for value conversion, then the associated user terminals of the user terminal are identified, the focus of attention between the user terminal and the associated user terminals is determined, and the fit between the update frequency and the change frequency of the focus of attention is determined based on the update frequency of the game and the change frequency of the focus of attention. Combined with the time difference of the iteration of the interactive behavior corresponding to the focus of attention, the update attention representation value of the user terminal for the game is determined, so as to adjust the game advertising delivery method for the user terminal. If the user group is a category with low potential for value conversion, then game ads will be randomly delivered on the vertical platform at a unit frequency.

[0013] Furthermore, the process of identifying the associated user terminals of the aforementioned user terminal includes: Acquire interaction data between the user client and other user clients; Other user terminals that meet the interaction association conditions will be designated as interactive user terminals. The associated user terminal is determined based on the overlapping characteristics of the behavioral trajectories of the interactive user terminal and the user terminal. The interaction association conditions include the frequency of interaction with the user terminal within a predetermined period being greater than the interaction frequency threshold and the number of interaction topics being greater than the number of interaction topics threshold. The overlapping features include the number of overlapping interactive themes and the average browsing time of overlapping interactive themes.

[0014] Further, determining the associated user terminal includes: If the number of overlaps between any interactive user terminal and the corresponding operation node of the user terminal is greater than the overlap threshold, and the average browsing time of the overlapping operation nodes is greater than the average browsing time threshold, then the interactive user terminal is identified as the associated user terminal.

[0015] Furthermore, the process of determining the focused topics of interest between the user terminal and the associated user terminal includes: Extract the maximum percentage of interactions between user terminals and related user terminals on several overlapping interactive topics, as well as the positiveness of interactive discussions; If the maximum percentage of interactions for any overlapping interactive topic is greater than the maximum percentage threshold for interactions, and the positive degree of interaction discussion is greater than the positive degree threshold for interaction discussion, then the overlapping interactive topic is determined as the topic of focus.

[0016] Further, the process of determining the fit between the update frequency and the replacement frequency includes: The update time node of the game update and the interaction time node of the user's interaction behavior in response to the game update are locked, and the average time difference between the update time node and the interaction time node is determined. The degree of fit is the sum of the ratio of the overlap between the interactive content corresponding to the interactive behavior and the update content corresponding to the game update to the overlap threshold, and the ratio of the average time difference threshold to the average time difference.

[0017] Further, determining the user's attention value to game updates includes: The ratio of the fit between update frequency and replacement frequency to the fit threshold is used as the first update focus feature; The ratio of the iteration time difference threshold to the iteration time difference of the corresponding interactive behavior for the focused topic is used as the second update attention feature. The sum of the first update attention feature and the second update attention feature is determined as the update attention representation value.

[0018] Furthermore, the method for delivering game advertisements to the user's end is adjusted, including: If the user's attention value for game updates is greater than or equal to the update attention threshold, then the game's real-time update content will be displayed on the vertical platform. If the user's attention value for game updates is less than the update attention threshold, then game ads will be randomly displayed on the vertical platform at a unit frequency.

[0019] Compared to existing technologies, this invention extracts corresponding daily engagement characteristics based on user engagement data for games. Combining these daily engagement characteristics with the proportion of time users spend browsing game-related content on vertical platforms, it analyzes conversion value representation parameters to categorize users into value conversion potential categories. Based on these value conversion potential categories, it adaptively adjusts and analyzes game advertising placement for users. This invention precisely targets user interests through a dual mechanism, making game advertising content more aligned with users' actual interests, improving user acceptance and experience, reducing ineffective interference, and matching the game's update rhythm to ensure timely advertising that keeps pace with demand and meets dynamic changes in user interests. Simultaneously, it improves advertising conversion efficiency, reduces placement costs, enhances the quality of the advertising ecosystem on vertical platforms, balances advertising monetization revenue with user experience protection, and strengthens user stickiness.

[0020] In particular, this invention, through multi-dimensional and precise user engagement data fusion analysis, characterizes the true depth of user engagement in the game. This includes: quantifying the actual quality and effective participation of users within the game by the ratio of active time to online time within a predetermined period, which helps distinguish between effectively active users and inactive users. The time interval deviation of online status switching reflects the stability and habitual regularity of user behavior, determining whether a user is a "loyal and stable user" of the game. Furthermore, by overlaying the proportion of user browsing time for game themes on vertical platforms, a dual screening of engagement depth and interest matching is achieved, identifying users with high conversion potential. The proportion of browsing time for game themes on vertical platforms represents the user's interest and focus on specific game themes. Therefore, this invention, through precise insight into user game needs, analyzes the conversion value representation parameters of users to characterize the probability of user response to game advertisements, conversion likelihood, and commercial value potential. This provides clear data support for subsequent classification of user value conversion potential categories, improving the accuracy of advertising, reducing resource waste, and optimizing user experience.

[0021] In particular, this invention achieves efficient identification of associated user terminals and focused topics by employing multi-condition filtering and precise judgment. Through a two-layer filtering mechanism constructed using interactive association conditions and overlapping features, it can accurately eliminate unrelated user terminals with occasional interactions and identify core associated user terminals with highly consistent behavioral trajectories, avoiding misjudgments of user groups due to invalid social data. The interaction frequency within a predetermined period reflects the tightness of the social connection and the level of interaction activity between user terminals and interacting user terminals. A higher interaction frequency within the predetermined period indicates more frequent contact and a greater likelihood of non-occasional interaction, serving as a fundamental indicator for determining whether a stable social connection exists between user terminals. Using this as the first threshold for screening associated user terminals quickly eliminates users with occasional interactions, reducing redundancy in subsequent data processing and improving the efficiency of associated user terminal identification. The corresponding number of interactive topics quantifies the breadth of interaction and shared topic foundation between user terminals and interacting user terminals. This feature effectively overcomes the limitations of relying solely on interaction frequency, avoiding the misclassification of users who "interact frequently on a single, accidental topic but have no other shared interests" as related users. This further improves the accuracy of related user screening, ensuring that the selected related users share a multi-dimensional common interest base. Furthermore, determining the number of overlapping interactive topics reflects the degree of overlap and focus of interests between users and interactive users. Correspondingly, the average browsing time for overlapping interactive topics further quantifies the level of focus and depth of engagement of users and interactive users on shared interest topics. By eliminating users who "have overlapping interactive topics but low focus," it ensures that related users are genuinely interested in shared topics. This indicator upgrades the judgment of related user interests from "whether there is overlap" to "whether the overlap is deep," further enhancing the conversion potential of advertising. Therefore, this invention, based on the aforementioned four features, progressively and synergistically constructs a precise related user screening system of "first screening frequency and breadth, then core overlap and depth." By analyzing the frequency and number of interactions within a predetermined period, we can initially identify user groups with stable interactions and a wide range of shared topics. Then, by analyzing the number of overlapping interactions and the average browsing time for overlapping interactions, we can precisely identify relevant user groups with highly compatible interests and sufficient engagement.

[0022] In particular, this invention mines stable and focused core interest topics, improving the accuracy of interest identification on the user end. It characterizes the user's true level of attention to overlapping interactive topics from two dimensions: interaction popularity and attitude tendency. The maximum percentage of interactions between the user end and related user ends regarding several overlapping interactive topics reflects the coreness and focus intensity of a particular overlapping interactive topic across all interactions between the user end and related user ends. Correspondingly, such stable and highly focused core topics are often the areas where users are most likely to generate ad responses and have the highest conversion intentions, and their underlying commercial value potential is also more prominent. The invention also positively measures the subjective preference and acceptance attitude of the user end and related user ends towards overlapping interactive topics by combining interactive discussions. This is derived through sentiment analysis of the interactive content of both parties regarding overlapping interactive topics. It effectively excludes topics with "high-frequency interaction but negative evaluations," ensuring that the mined focused topics are content that both parties genuinely like. This approach guarantees both high interactivity of overlapping interactive topics and a positive attitude from the user end towards those topics. Compared to single-dimensional interest mining, it better reflects the user's true and stable core interests, avoiding misjudgments of interests caused by short-term random browsing. Therefore, this invention focuses on topics that are of common interest and recognition to both the user and related users. By placing advertisements based on this, the matching degree between the advertisement content and the user's interests can be greatly improved, the user's acceptance of the advertisement can be enhanced, and the advertisement conversion effect can be improved.

[0023] In particular, this invention integrates two core elements—the overlap between game update content and interactive content, and the time difference in the iteration of interactive behaviors corresponding to updates on focused themes—to quantify the matching rhythm between game updates and changes in user interests, thereby improving the timeliness of game advertising. The alignment between update frequency and change frequency reflects the degree of synchronization between the game update rhythm and the changing rhythm of users' focused themes, as well as the quality of user response to game update content, thus measuring the degree of matching between game dynamics and user needs. This ensures that advertising content matches the core information of game updates while guaranteeing that the timing of ad placement aligns with the user's attention rhythm. The time difference in the iteration of interactive behaviors corresponding to focused themes represents the time interval between user interactions on these themes, quantifying the persistence of user interest and the stability of behavioral dynamics. Furthermore, a narrowing time difference indicates increased user interest in the theme, and increasingly frequent interactions are a direct reflection of strong user interest. Therefore, this invention determines the user's attention level to game updates by measuring the user's response sensitivity, interest level, and potential response and conversion potential to update-related advertisements. This provides data support for subsequent adjustments to game advertising delivery methods for users. This invention achieves accurate assessment of user attention to game updates, thereby providing a basis for adjusting advertising delivery methods. Attached Figure Description

[0024] Figure 1 A schematic diagram illustrating the steps of a user profile-based game advertising delivery method according to an embodiment of the invention; Figure 2 A logic decision diagram for classifying the value conversion potential categories of the user end in the embodiments of the invention; Figure 3 A logic decision diagram for determining the associated user terminal in the embodiments of the invention; Figure 4 A logic diagram for determining the focus of attention between a user terminal and associated user terminals in an embodiment of the invention. Detailed Implementation

[0025] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0026] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0027] Please see Figure 1 The diagram illustrates the steps of a user-profile-based game advertising delivery method according to an embodiment of the present invention. The user-profile-based game advertising delivery method according to an embodiment of the present invention includes: Step S1: Based on the user's investment data in the game, extract the corresponding daily investment features, which include the ratio of active time to online time within a predetermined period and the deviation value of the time interval between online status switching. Step S2: Combining the daily input characteristics and the proportion of browsing time of game-related content on the vertical platform, analyze the conversion value representation parameters of the user terminal to classify the value conversion potential categories of the user terminal. Step S3, based on the value conversion potential category, performs adjustment analysis on the game advertising placement on the user end, including, Identify the associated user terminals of the user terminal, determine the focus of attention between the user terminal and the associated user terminals, determine the fit between the update frequency and the change frequency of the focus of attention based on the update frequency of the game and the change frequency, and determine the update attention characterization value of the user terminal for the game by combining the time difference of the interaction behavior corresponding to the focus of attention. In order to adjust the game advertising delivery method for the user terminal. Game ads are randomly displayed on vertical platforms at unit frequencies.

[0028] Specifically, the vertical platform refers to a digital network platform built for the gaming industry, focusing on the distribution of game-related content, user interaction, and discussion of themes, serving user groups with a preference for games, including but not limited to official game forums, player communities, and game review websites.

[0029] Specifically, in some possible implementations, game ads can be replaced with other ads and pushed to the corresponding vertical platforms.

[0030] Specifically, the input data includes daily input characteristics, the percentage of time users spend browsing game-related content on vertical platforms, and interaction data between users and other users. The interaction data includes the frequency of interaction with the user terminal within a predetermined period and the number of corresponding interaction topics, overlapping characteristics, the number of overlapping operation nodes between the user terminal and the user terminal, the average browsing time of overlapping operation nodes, the maximum percentage of interaction times between the user terminal and related user terminals for several overlapping interaction topics and the positiveness of interaction discussions, the average time difference between update time nodes and interaction time nodes, the degree of overlap between the interaction content corresponding to the interaction behavior and the update content corresponding to the game update, and the time difference of the iteration of the interaction behavior corresponding to the focus of attention.

[0031] Specifically, relevant input data is collected through multiple channels, including user-side game behavior records, vertical platform browsing logs, and user-side interaction data records. Before acquiring the game behavior records, the vertical platform browsing logs, and the interaction data records, the user must be clearly informed of the purpose of the data and the user's consent must be obtained. The data will only be used for analysis related to the precise targeting of game advertisements, which will not be elaborated further here.

[0032] Specifically, the ratio of active time to online time within the predetermined period refers to the cumulative time during which the user's client generates effective interactive behaviors in the game. Effective interactive behaviors include, but are not limited to, completing game tasks, participating in competitive battles, engaging in player chat interactions, and triggering game function operations. Online time refers to the cumulative total time from when the user's client account logs into the game system until the account actively logs out or the system determines that the user is offline.

[0033] Understandably, the gaming industry typically develops operational plans on a weekly basis, such as updating events, releasing weekly tasks, and setting leaderboards. User gaming behavior also exhibits clear weekly patterns, such as fragmented logins on weekdays and concentrated long-term gaming sessions on weekends. Therefore, to fully cover all scenarios of user behavior, from light daily engagement to deep weekend experiences, and to avoid the randomness of daily data or the lag of monthly data (such as failing to capture changes in user interest), ensuring that the collected engagement data better reflects users' true gaming habits, the predetermined period is set to one week.

[0034] Specifically, the online status switching refers to the change in the login status of the user's account within the game system during the game's operation. The core includes the complete "online-offline" status switching, covering status changes triggered by user-initiated operations, such as active login and manual logout, as well as status changes automatically determined by the system, such as being kicked offline by the system due to timeout failure or passive offline due to network interruption.

[0035] Specifically, the process of analyzing the conversion value representation parameters of the user end includes: The sum of the ratio of active time to online time within a predetermined period to the time percentage threshold and the ratio of the time interval deviation threshold to the time interval deviation value of online status switching is used as the first conversion value feature. The ratio of the percentage of time users spend browsing game-related content on vertical platforms to the threshold percentage of time users spend browsing game-related content is used as the second conversion value feature. The first conversion value feature and the second conversion value feature are weighted and summed to form the conversion value representation parameter.

[0036] Specifically, in practice, ad conversion hinges on users' sustained willingness to engage with the game. Users with high active time percentages are more likely to interact effectively while online rather than being offline, demonstrating strong engagement. Users with small deviations in online status switching intervals have stable gaming habits and represent the core user group. These users pay more attention to game-related ads, and their conversion rates (clicks, downloads, payments, etc.) are significantly higher than those of users with low engagement and unstable behavior. Therefore, the primary conversion value characteristic focuses on the quality and stability of users' actual engagement within the game, directly reflecting their "loyalty, immersion, and effective participation," which is the core foundation for ad conversion. The percentage of time users spend browsing game-related content on vertical platforms further filters out users with high interest alignment from the existing potential conversion group. Only when users have established stable gaming engagement can interest-based matching significantly improve conversion efficiency. Therefore, the first conversion value characteristic, calculated based on daily input characteristics—namely, the ratio of active time to online time within a predetermined period and the deviation value of the time interval between online status transitions—is given a higher weighting coefficient, set to 0.6. Correspondingly, the weighting coefficient for the second conversion value characteristic, calculated based on the proportion of user browsing time for game-related content on vertical platforms, is set to 0.4.

[0037] In this embodiment, the purpose of setting the time duration percentage threshold, the time interval deviation threshold, and the game-themed browsing time percentage threshold is to characterize situations where the user's effective investment depth in the game is high, the user's gaming habits have certain regularity, and the user's response probability to game advertisements is high, with a high conversion possibility. By acquiring historical user engagement data for the same game, and retrieving historical data on the ratio of active time to online time within a predetermined period, historical data on the time interval deviation of online status switching, and historical data on the ratio of user browsing time for game-related topics on vertical platforms, the average duration ratio, average time interval deviation, and average game-related topic browsing time ratio are calculated and used as baseline values ​​under normal circumstances. Based on the purpose of setting the above three thresholds, the duration ratio threshold is determined as the product of the average duration ratio and a first deviation coefficient; the time interval deviation threshold is determined as the product of the time interval deviation and a second deviation coefficient; and the game-related topic browsing time ratio threshold is determined as the product of the average game-related topic browsing time ratio and a third deviation coefficient. Specifically, the first deviation coefficient is selected within the range [1.4, 1.6], preferably 1.4; the second deviation coefficient is selected within the range [0.9, 0.95], preferably 0.9; and the third deviation coefficient is selected within the range [1.3, 1.5], preferably 1.3.

[0038] Specifically, the deviation value of the online state switching time interval is determined by calculating the standard deviation of the time intervals between several state switching events on the user's end within a predetermined period. A smaller standard deviation indicates a more uniform online state switching interval on the user's end, more stable behavior, and a higher degree of regularity in its game usage habits.

[0039] Specifically, this invention uses multi-dimensional and precise user engagement data fusion analysis to characterize the true depth of user engagement in the game. This includes: quantifying the actual quality of user engagement and effective participation in the game by measuring the ratio of active time to online time within a predetermined period. This helps distinguish between effectively active users and inactive users. The higher the ratio, the stronger the user's effective participation while online, and the higher their immersion and stickiness in the game. For example, if a user spends 10 hours online per week, with 8 hours of active time (80%), it indicates that this user is primarily engaged in the game and has high conversion potential. Conversely, if another user also spends 10 hours online, but only 2 hours (20%) are active, they are likely inactive, with low actual engagement and relatively low conversion value.

[0040] The deviation value of the time interval between online status switching reflects the stability and regularity of user behavior when playing games, determining whether a user is a "loyal and stable user." A smaller deviation value indicates a regular online / offline gaming time pattern and stable gaming habits; these users have higher loyalty to the game and more consistent attention to game updates, promotional activities, and other advertising content. Conversely, a larger deviation value indicates more random user behavior, suggesting occasional, sporadic users with less predictable advertising responses and relatively fluctuating conversion value. Furthermore, by combining this with the percentage of time users spend browsing game-related topics on vertical platforms, a dual screening of engagement depth and interest matching is achieved, identifying users with high conversion potential. The percentage of time spent browsing game-related topics on vertical platforms represents the user's interest and focus on specific game topics. A higher percentage of time spent browsing game-related topics indicates a stronger user interest and more concentrated attention on that topic. For example, if 70% of a user's browsing time on a vertical platform is focused on gaming, it indicates that their core interest is in gaming, and their acceptance and conversion potential for ads in this genre are relatively higher. Conversely, if a user's browsing time is spread across multiple genres without a clearly dominant category, it indicates broad interests, and the accuracy of targeted advertising needs to be further assessed using other indicators. Therefore, this invention, through precise insights into user gaming needs, analyzes the conversion value representation parameters of users to characterize the probability of user response to gaming ads, conversion likelihood, and commercial value potential. This provides clear data support for subsequently categorizing users' value conversion potential, improving ad targeting accuracy, reducing resource waste, and optimizing user experience.

[0041] Specifically, please refer to Figure 2 As shown, it is a logical decision diagram for classifying the value conversion potential categories of the user terminal according to an embodiment of the present invention. The classification of the value conversion potential categories of the user terminal includes: If the conversion value representation parameter of the user terminal is greater than or equal to the conversion value representation parameter threshold, then the user terminal is classified as a high-potential value conversion category. If the conversion value representation parameter of the user terminal is less than the conversion value representation parameter threshold, then the user terminal is classified as a low-potential value conversion category.

[0042] The conversion value representation parameter threshold is predetermined. It is determined by assuming that the proportion of active time to online time within a predetermined period is equal to the time proportion threshold, the time interval deviation threshold is equal to the time interval deviation value of online status switching, and the proportion of user terminal browsing time of game-related content on the vertical platform is equal to the game-related content browsing time proportion threshold.

[0043] Specifically, the clear categorization of conversion value potential categories in this invention guides the allocation of advertising resources. It prioritizes high-quality advertising resources and refined targeting strategies for users with high conversion value, ensuring effective campaigns. It also reduces ineffective targeting of users with low conversion potential by employing a low-cost, random targeting model for these users, controlling overall costs. This precise stratification makes advertising resource allocation more rational, avoiding resource waste caused by a "one-size-fits-all" approach and improving overall advertising resource utilization.

[0044] Specifically, the analysis and adjustment of game advertising on the user's end includes: If the user terminal is a category with high potential for value conversion, then the associated user terminals of the user terminal are identified, the focus of attention between the user terminal and the associated user terminals is determined, and the fit between the update frequency and the change frequency of the focus of attention is determined based on the update frequency of the game and the change frequency of the focus of attention. Combined with the time difference of the iteration of the interactive behavior corresponding to the focus of attention, the update attention representation value of the user terminal for the game is determined, so as to adjust the game advertising delivery method for the user terminal. If the user group is a category with low potential for value conversion, then game ads will be randomly delivered on the vertical platform at a unit frequency.

[0045] Specifically, the process of identifying the associated user terminals of the aforementioned user terminal includes: Acquire interaction data between the user client and other user clients; Other user terminals that meet the interaction association conditions will be designated as interactive user terminals. The associated user terminal is determined based on the overlapping characteristics of the behavioral trajectories of the interactive user terminal and the user terminal. The interaction association conditions include the frequency of interaction with the user terminal within a predetermined period being greater than the interaction frequency threshold and the number of interaction topics being greater than the number of interaction topics threshold. The overlapping features include the number of overlapping interactive themes and the average browsing time of overlapping interactive themes.

[0046] In this embodiment, the purpose of setting the interaction frequency threshold is to characterize the high degree of social connection and high level of interaction activity between user terminals and other user terminals. The purpose of setting the interaction topic quantity threshold is to characterize the breadth and depth of interaction between user terminals and other user terminals, and the solid foundation of common topics. By acquiring historical interaction data of user terminals for the same game, calling historical interaction frequency data and corresponding historical interaction topic quantity data within a predetermined period, the mean interaction frequency and the mean interaction topic quantity are calculated and used as the benchmark values ​​under normal circumstances. Based on the purpose of setting the above two thresholds, the interaction frequency threshold is determined as the product of the mean interaction frequency and the frequency deviation coefficient, and the interaction topic quantity threshold is determined as the product of the mean interaction topic quantity and the quantity deviation coefficient. The frequency deviation coefficient is selected within the interval [1.4, 1.5], preferably 1.4 in the implementation, and the quantity deviation coefficient is selected within the interval [1.4, 1.6], preferably 1.4 in the implementation.

[0047] Specifically, please refer to Figure 3 As shown, this is a logic decision diagram for determining the associated user terminal in an embodiment of the present invention. Determining the associated user terminal includes: If the number of overlaps between any interactive user terminal and the corresponding operation node of the user terminal is greater than the overlap threshold, and the average browsing time of the overlapping operation nodes is greater than the average browsing time threshold, then the interactive user terminal is identified as the associated user terminal.

[0048] In this embodiment, the purpose of setting the overlap quantity threshold is to characterize the situation where the degree of overlap of interests between the user terminal and the interactive user terminal is high, and the focus range is relatively close. The purpose of setting the average browsing time threshold is to characterize the situation where the user terminal and the interactive user terminal have a high degree of focus and deep engagement on common interest topics. By obtaining the historical interaction data of the user terminal for the same game, calling the historical data of the overlap quantity of the corresponding operation nodes of the interactive user terminal and the user terminal, and the historical data of the average browsing time of the overlapping operation nodes, the mean of the overlap quantity and the mean of the average browsing time are calculated, and the corresponding values ​​are used as the benchmark values ​​under normal circumstances. Based on the purpose of setting the above two thresholds, the overlap quantity threshold is determined to be the product of the mean overlap quantity and the overlap deviation coefficient, and the average browsing time threshold is determined to be the product of the mean average browsing time and the browsing deviation coefficient. The overlap deviation coefficient is selected in the interval [1.3, 1.5], preferably 1.3 in the implementation, and the browsing deviation coefficient is selected in the interval [1.4, 1.6], preferably 1.4 in the implementation.

[0049] Specifically, this invention achieves efficient identification of associated user accounts and focused topics by employing multi-condition filtering and precise judgment. Through a two-layer filtering mechanism constructed using interaction association conditions and overlapping features, it can accurately eliminate unrelated user accounts with occasional interactions and identify core associated user accounts with highly consistent behavioral trajectories, avoiding misjudgments of user groups due to invalid social data. The interaction frequency within a predetermined period reflects the strength of the social connection and the level of interaction activity between user accounts and interacting user accounts. A higher interaction frequency within the predetermined period indicates more frequent contact and a greater likelihood of non-occasional interaction, serving as a fundamental indicator for determining whether a stable social connection exists between user accounts. For example, user accounts that interact more than 10 times per week have a significantly stronger social connection than those that interact only 3 times per week. Using this as the first threshold for filtering associated user accounts can quickly eliminate users with occasional interactions, reducing redundancy in subsequent data processing and improving the efficiency of associated user account identification.

[0050] The number of interactive topics quantifies the breadth of interaction and shared topics between the user client and the interacting user client. A higher number of interactive topics indicates a wider overlap in interests and a greater potential for deep connection. This feature effectively compensates for the limitations of relying solely on interaction frequency, avoiding misclassifying interacting user clients who "only interact frequently on a particular occasional topic but have no other shared topics" as related user clients, further improving the accuracy of related user client selection and ensuring that the selected related user clients possess a multi-dimensional shared interest base. Furthermore, the number of overlapping interactive topics reflects the degree of overlap and focus of interests between the user client and the interacting user client; a higher overlap indicates a more aligned core interest and a higher similarity in behavioral patterns. Correspondingly, the average browsing time for overlapping interactive topics further quantifies the focus and depth of engagement of the user client and the interacting user client on shared interest topics. A longer average browsing time indicates a stronger interest and greater willingness to engage in the overlapping interactive topics. By eliminating user clients with "overlapping interactive topics but low focus," it ensures that related user clients are genuinely interested in shared topics. This metric upgrades the judgment of related user interests from "whether there is overlap" to "whether the overlap is deep," further enhancing the conversion potential of ad placement. Therefore, this invention constructs a precise screening system for related users based on the aforementioned four features, which are progressively integrated and mutually reinforcing. First, frequency and breadth are screened; then, overlap and depth are assessed. By analyzing the frequency and number of interaction topics within a predetermined period, users with stable interactions and broad shared topics are initially identified. Then, by analyzing the number of overlapping interaction topics and the average browsing time of overlapping topics, highly compatible users with sufficient engagement are precisely identified.

[0051] Specifically, please refer to Figure 4As shown, this is a logic diagram for determining the focus of attention between a user terminal and an associated user terminal according to an embodiment of the present invention. The process of determining the focus of attention between the user terminal and the associated user terminal includes: Extract the maximum percentage of interactions between user terminals and related user terminals on several overlapping interactive topics, as well as the positiveness of interactive discussions; If the maximum percentage of interactions for any overlapping interactive topic is greater than the maximum percentage threshold for interactions, and the positive degree of interaction discussion is greater than the positive degree threshold for interaction discussion, then the overlapping interactive topic is determined as the topic of focus.

[0052] In this embodiment, the purpose of setting the maximum percentage threshold for interaction frequency is to characterize the situation where a certain overlapping interactive theme has a high degree of coreness and focus in all interactions between the user and related user ends. The purpose of setting the positive interaction discussion threshold is to characterize the situation where the user and related user ends have a high degree of subjective preference and recognition for overlapping interactive themes. By obtaining historical interaction data of the user end for the same game, calling the historical data of the maximum percentage of interaction frequency for several overlapping interactive themes between the user end and related user ends, and the historical data of the positive interaction discussion, the average maximum percentage of interaction frequency and the average positive interaction discussion are calculated and used as the benchmark value under normal circumstances. Based on the purpose of setting the above two thresholds, the maximum percentage threshold for interaction frequency is determined as the product of the average maximum percentage of interaction frequency and the first offset coefficient, and the positive interaction discussion threshold is determined as the product of the average positive interaction discussion and the second offset coefficient. The first offset coefficient is selected in the interval [1.5, 1.6], preferably 1.5 in the implementation, and the second offset coefficient is selected in the interval [1.3, 1.4], preferably 1.3 in the implementation.

[0053] Specifically, the positive aspect of the interactive discussion refers to the proportion of positive emotions, including but not limited to recognition, liking, and anticipation, contained in the overall emotions when users interact with each other on a vertical platform around overlapping interactive themes related to the game.

[0054] This process involves processing the interactive text into a fixed-length sequence and converting it into a low-dimensional semantic vector using pre-trained word vectors from the game domain, such as the Word2Vec model trained on game forum corpora. A deep learning module adapted for text sentiment analysis, such as LSTM or BERT, is then used as the input for training. The model's output layer is set to a probability output structure; for example, after fine-tuning the BERT model for game sentiment analysis scenarios, its output text is determined to have a confidence level indicating positive sentiment. This confidence level is then used as the positive sentiment level of the interactive discussion.

[0055] Specifically, this invention identifies stable and focused core interest topics to improve the accuracy of interest identification on the user end. It characterizes the user's true level of attention to overlapping interactive topics from two dimensions: interaction popularity and attitude tendency. The maximum percentage of interactions between a user end and related user ends regarding several overlapping topics reflects the core nature and focus of a particular overlapping topic across all interactions between the user end and related user ends. A high maximum percentage indicates a more stable and focused shared interest between the user end and related user ends, rather than a short-term, randomly fluctuating interest. Most interactions revolve around this topic, indicating that it aligns with both parties' long-term stable interest preferences. Correspondingly, such stable and highly focused core topics are often the areas where users are most likely to generate ad responses and have the highest conversion rates, and their underlying commercial value potential is also more prominent. The invention also combines positive interaction discussions to quantify the subjective liking and acceptance of overlapping interactive topics by both users and related user ends. This is achieved through sentiment analysis of interactive content related to overlapping topics, such as topic comments and content replies. The higher the positiveness of the interaction discussions, the more positive evaluations and stronger the acceptance of the overlapping interactive topic by both parties. This method effectively excludes topics with "high-frequency interaction but negative reviews," such as frequent complaints about the shortcomings of a particular topic, ensuring that the selected topics are content that both parties genuinely enjoy. It guarantees both high interactivity of overlapping topics and a positive user attitude towards them. Compared to single-dimensional interest mining, this method better reflects users' true and stable core interests, avoiding misjudgments of interests caused by short-term random browsing. Therefore, this invention focuses on topics that are content that both the user and related users are interested in and agree upon. Using this as a basis for ad placement significantly improves the match between ad content and user interests, enhances user acceptance of ads, and ultimately improves ad conversion rates.

[0056] Specifically, the process of determining the fit between the update frequency and the replacement frequency includes: The update time node of the game update and the interaction time node of the user's interaction behavior in response to the game update are locked, and the average time difference between the update time node and the interaction time node is determined. The degree of fit is the sum of the ratio of the overlap between the interactive content corresponding to the interactive behavior and the update content corresponding to the game update to the overlap threshold, and the ratio of the average time difference threshold to the average time difference.

[0057] In this embodiment, the purpose of setting the overlap threshold is to characterize situations where the user's interactive content and the game's updated content have a high degree of matching. The purpose of setting the average time difference threshold is to characterize situations where the user's response to the game update is likely to be converted. By acquiring historical interaction data of the user for the same game, historical data on the overlap between the interactive content corresponding to the interactive behavior and the updated content corresponding to the game update, as well as historical data on the average time difference between the update time node and the interaction time node, the mean overlap and the mean average time difference are calculated and used as the baseline values ​​under normal circumstances. Based on the purpose of setting the above two thresholds, the overlap threshold is determined to be the product of the mean overlap and the overlap deviation coefficient, and the mean time difference threshold is determined to be the product of the mean average time difference and the mean deviation coefficient. The overlap deviation coefficient is selected within the interval [1.3, 1.5], preferably 1.3 in practice, and the mean deviation coefficient is selected within the interval [0.9, 0.95], preferably 0.9 in practice.

[0058] Specifically, the degree of overlap between the interactive content corresponding to the interactive behavior and the update content corresponding to the game update can be determined by any existing device, algorithm or system capable of detecting and quantifying the content overlap, which will not be elaborated further here.

[0059] Specifically, determining the user's attention value for game updates includes: The ratio of the fit between update frequency and replacement frequency to the fit threshold is used as the first update focus feature; The ratio of the iteration time difference threshold to the iteration time difference of the corresponding interactive behavior for the focused topic is used as the second update attention feature. The sum of the first update attention feature and the second update attention feature is determined as the update attention representation value.

[0060] In this embodiment, the purpose of setting the fit threshold is to characterize a high degree of synchronization between the game update rhythm and the user's focus on the changing interest rhythm of the theme. The purpose of setting the iteration time difference threshold is to characterize a long duration of user interest in the focus theme and a high degree of dynamic stability of behavior. By acquiring the user's historical interaction data for the same game, calling the historical data of fit between update frequency and change frequency, and the historical data of iteration time difference for the corresponding interactive behavior of the focus theme, and using the corresponding values ​​as the baseline under normal circumstances, based on the purpose of setting the above two thresholds, the fit threshold is determined as the product of the fit mean and the fit deviation coefficient, and the iteration time difference threshold is determined as the product of the iteration time difference mean and the iteration deviation coefficient. The fit deviation coefficient is selected in the interval [1.3, 1.], preferably 1.3 in the implementation, and the iteration deviation coefficient is selected in the interval [0.9, 0.95], preferably 0.9 in the implementation.

[0061] Specifically, this invention integrates two core elements: the overlap between game update content and interactive content, and the time difference in the iteration of interactive behaviors corresponding to updates on focused themes, to quantify and match the rhythm of game updates with changes in user interests, thereby improving the timeliness of game advertising. The alignment between update frequency and change frequency reflects the degree of compatibility and synchronization between the game update rhythm and the rhythm of changes in user interests regarding focused themes, as well as the quality of user response to game update content. This measures the degree of matching between game dynamics and user needs; a high alignment indicates that the game update accurately targets the changing nodes of user interests. While ensuring that the ad content matches the core information of the game update, it also ensures that the timing of the ad placement aligns with the user's attention rhythm. For example, if a game updates a new gameplay mode on a focused theme, and users quickly engage in interactive discussions with highly relevant content, the alignment is high, and placing the corresponding ad at this time can accurately hit the user's need for novelty. The time difference in the iteration of interactive behaviors corresponding to focused themes is the time interval between user interactions on focused themes, quantifying the persistence of user interest in focused themes and the stability of behavioral dynamics. The smaller the update time difference, the more regular and consistent the user's interactive behavior regarding this topic, indicating a sustained and stable interest in it. Furthermore, a smaller update time difference indicates increased user attention to the topic, and more frequent interaction is a direct reflection of strong user interest. Therefore, this invention determines the user's update attention characteristic value for the game to represent the user's sensitivity to game updates, the intensity of interest, and the potential response and conversion potential to update-related advertisements, providing data support for subsequent adjustments to game advertising delivery methods. This invention achieves accurate assessment of user attention to game updates, thereby providing a basis for adjusting advertising delivery methods.

[0062] Specifically, the method for delivering game ads to the user's device is adjusted, including: If the user's attention value for game updates is greater than or equal to the update attention threshold, then the game's real-time update content will be displayed on the vertical platform. If the user's attention value for game updates is less than the update attention threshold, then game ads will be randomly displayed on the vertical platform at a unit frequency.

[0063] The updated attention representation threshold is predetermined. The updated attention representation value is determined by calculating the following conditions: the degree of fit between the update frequency and the replacement frequency is equal to the degree of fit threshold, and the time difference threshold is equal to the time difference of the time difference of the interaction behavior corresponding to the focused attention topic.

[0064] If the user profile-based game advertising delivery method of the present invention is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the various embodiments of the present invention. The aforementioned storage medium includes various media that can store program code, such as USB flash drives, mobile hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0065] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles 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 all fall within the scope of protection of the present invention.

Claims

1. A method for game advertising based on user profiles, characterized in that, include: Based on user data on game engagement, corresponding daily engagement features are extracted. These features include the ratio of active time to online time within a predetermined period and the deviation value of the time interval between online status switching. By combining the daily input characteristics and the proportion of browsing time of game-related content on the vertical platform, the conversion value representation parameters of the user terminal are analyzed to classify the value conversion potential categories of the user terminal. Based on the aforementioned value conversion potential categories, adjustments and analyses are made to the game advertising placement on the user's end, including: Identify the associated user terminals of the user terminal, determine the focus of attention between the user terminal and the associated user terminals, determine the fit between the update frequency and the change frequency of the focus of attention based on the update frequency of the game and the change frequency, and determine the update attention characterization value of the user terminal for the game by combining the time difference of the interaction behavior corresponding to the focus of attention. In order to adjust the game advertising delivery method for the user terminal. Game ads are randomly displayed on vertical platforms at unit frequencies.

2. The game advertising delivery method based on user profiles according to claim 1, characterized in that, The process of analyzing the conversion value representation parameters of the user end includes: The sum of the ratio of active time to online time within a predetermined period to the time percentage threshold and the ratio of the time interval deviation threshold to the time interval deviation value of online status switching is used as the first conversion value feature. The ratio of the percentage of time users spend browsing game-related content on vertical platforms to the threshold percentage of time users spend browsing game-related content is used as the second conversion value feature. The first conversion value feature and the second conversion value feature are weighted and summed to form the conversion value representation parameter.

3. The game advertising delivery method based on user profiles according to claim 2, characterized in that, The value conversion potential categories of the aforementioned user terminals are divided into: If the conversion value representation parameter of the user terminal is greater than or equal to the conversion value representation parameter threshold, then the user terminal is classified as a high-potential value conversion category. If the conversion value representation parameter of the user terminal is less than the conversion value representation parameter threshold, then the user terminal is classified as a low-potential value conversion category.

4. The game advertising delivery method based on user profiles according to claim 3, characterized in that, The analysis and adjustment of game advertising on the user's end includes: If the user terminal is a category with high potential for value conversion, then the associated user terminals of the user terminal are identified, the focus of attention between the user terminal and the associated user terminals is determined, and the fit between the update frequency and the change frequency of the focus of attention is determined based on the update frequency of the game and the change frequency of the focus of attention. Combined with the time difference of the iteration of the interactive behavior corresponding to the focus of attention, the update attention representation value of the user terminal for the game is determined, so as to adjust the game advertising delivery method for the user terminal. If the user group is a category with low potential for value conversion, then game ads will be randomly delivered on the vertical platform at a unit frequency.

5. The game advertising delivery method based on user profiles according to claim 1, characterized in that, The process of identifying the associated user terminals of the aforementioned user terminal includes: Acquire interaction data between the user client and other user clients; Other user terminals that meet the interaction association conditions will be designated as interactive user terminals. The associated user terminal is determined based on the overlapping characteristics of the behavioral trajectories of the interactive user terminal and the user terminal. The interaction association conditions include the frequency of interaction with the user terminal within a predetermined period being greater than the interaction frequency threshold and the number of interaction topics being greater than the number of interaction topics threshold. The overlapping features include the number of overlapping interactive themes and the average browsing time of overlapping interactive themes.

6. The game advertising delivery method based on user profiles according to claim 5, characterized in that, Determining the associated user terminal includes: If the number of overlaps between any interactive user terminal and the corresponding operation node of the user terminal is greater than the overlap threshold, and the average browsing time of the overlapping operation nodes is greater than the average browsing time threshold, then the interactive user terminal is identified as the associated user terminal.

7. The game advertising delivery method based on user profiles according to claim 1, characterized in that, The process of determining the focus of attention between the user terminal and the associated user terminal includes: Extract the maximum percentage of interactions between user terminals and related user terminals on several overlapping interactive topics, as well as the positiveness of interactive discussions; If the maximum percentage of interactions for any overlapping interactive topic is greater than the maximum percentage threshold for interactions, and the positive degree of interaction discussion is greater than the positive degree threshold for interaction discussion, then the overlapping interactive topic is determined as the topic of focus.

8. The game advertising delivery method based on user profiles according to claim 1, characterized in that, The process of determining the fit between the update frequency and the replacement frequency includes: The update time node of the game update and the interaction time node of the user's interaction behavior in response to the game update are locked, and the average time difference between the update time node and the interaction time node is determined. The degree of fit is the sum of the ratio of the overlap between the interactive content corresponding to the interactive behavior and the update content corresponding to the game update to the overlap threshold, and the ratio of the average time difference threshold to the average time difference.

9. The game advertising delivery method based on user profiles according to claim 1, characterized in that, Determining the user's attention value to game updates includes: The ratio of the fit between update frequency and replacement frequency to the fit threshold is used as the first update focus feature; The ratio of the iteration time difference threshold to the iteration time difference of the corresponding interactive behavior for the focused topic is used as the second update attention feature. The sum of the first update attention feature and the second update attention feature is determined as the update attention representation value.

10. The game advertising delivery method based on user profiles according to claim 9, characterized in that, Adjusting the game advertising delivery method for the aforementioned user terminals, including: If the user's attention value for game updates is greater than or equal to the update attention threshold, then the game's real-time update content will be displayed on the vertical platform. If the user's attention value for game updates is less than the update attention threshold, then game ads will be randomly displayed on the vertical platform at a unit frequency.

Images